I
MANIPULATION OF POST-HARVEST PHYSIOLOGY IN BROCCOLI THROUGH AN
OPTIMISED AGROBACTERIUM RHIZOGENES-MEDIATED
TRANSFORMATION PROTOCOL
by
JAMES DAVID HIGGINS
A thesis submitted to the Faculty of Science of The University of Birmingham
for the degree of DOCTOR OF PHILOSOPHY
School of Biosciences The University of Birmingham
January 2002
University of Birmingham Research Archive
e-theses repository This unpublished thesis/dissertation is copyright of the author and/or third parties. The intellectual property rights of the author or third parties in respect of this work are as defined by The Copyright Designs and Patents Act 1988 or as modified by any successor legislation. Any use made of information contained in this thesis/dissertation must be in accordance with that legislation and must be properly acknowledged. Further distribution or reproduction in any format is prohibited without the permission of the copyright holder.
II
Abstract
The aim of this project was to down-regulate ACC oxidase (ACO) 1
and 2 and ACC synthase (ACS) in broccoli (Brassica oleracea var. italica) to
lengthen post-harvest shelf-life. The ACO 1 and 2 and ACS cDNAs of broccoli
were ligated into pSCV1.0 in sense and antisense orientations in relation to a
CaMV 35S promoter and nos terminator. They were electroporated into the
Agrobacterium rhizogenes co-integrate strain LBA 9402 pRi1855::GFP, and
used to co-transform GDDH33, a doubled haploid line derived from the
calabrese cultivar Green Duke. 150 transgenic hairy roots were identified by
GFP fluorescence, and 18 were regenerated into whole plants. Four of these
lines showed severe rol phenotype, which did not appear to be related to TL-
DNA insert copy number. The floral buds from T0 broccoli heads were
assayed for post-harvest production of ethylene and chlorophyll levels. T0
lines with ACC oxidase 1 and 2 constructs produced significantly less
ethylene than the control plants, and chlorophyll loss was significantly
reduced. A positive correlation between post-harvest bud ethylene production
and chlorophyll loss was described by the equation y= 0.2386x-23.041. There
were two copies of aco1 and aco2 in the Brassica oleracea genome of which
one was mapped for each to linkage group 1 and 8, respectively.
Key Words: broccoli, Agrobacterium rhizogenes, transformation, ACC oxidase, ACC synthase, ethylene, shelf-life
3
Acknowledgements
I would first like to thank BBSRC and HRI for financially supporting my
research over the past 3 years, and Horticulture Research International,
Wellesbourne, and the University of Birmingham for providing me the
opportunity to do this work.
At HRI, the glasshouse staff for watering my plants, and taking care of
aphid problems. For Linda Doyle and Sandy McClement for pollinating and
harvesting seeds from the T0 plants.
I would like to thank Graham King and his group, Graham Teakle, Guy Barker and Rowena Naylor for assisting my work with the mapping.
I would also like to thank the breeding/transformation group of Helen
Robinson, Noel Cogan, Lesley Griffiths, Jason Pole, and Liz Harvey for
making my time at HRI enjoyable.
On a more personal level, I would like to thank Anne Morton who helped
me with the Brassica DNA extractions in her own time, and Dez Barbara who
also assisted beyond the call of duty. This thesis would not have been
complete without their help. I would also like to thank Tony Roberts, Vinodh,
Alex and Emily for making their laboratory a pleasant environment.
My special thanks go to Ian Puddephat, Dez Barbara and John Newbury
who have supported me throughout the 3 years with scientific advice, moral
support and humour when necessary.
My last acknowledgement goes to Jane Byrne and Erik Kop who have
made the last 3 years fun.
4
Table of Contents 1.0 General Introduction .................................................................................. 1
1.1 Introduction ................................................................................................ 2
1.2 Genetic Improvement of Crop Plants ......................................................... 4
1.3 Transformation of Brassica ........................................................................ 6
1.4 Senescence of Broccoli ........................................................................... 12
1.5 Ethylene Biosynthesis .............................................................................. 13
1.5.1 ACC synthase ................................................................................... 14
1.5.2 ACC oxidase ..................................................................................... 14
1.6 Signalling ................................................................................................. 15
1.7 Antisense Technology.............................................................................. 19
1.8 Isolation of ACC genes ............................................................................ 21
1.9 Tissue specific expression of ACC oxidases in broccoli .......................... 23
1.10 Tomato fruit ripening and broccoli post-harvest senescence ................. 25
1.11 Project Aims........................................................................................... 27
2.0 Construction of Brassica oleracea ACC oxidase and ACC synthase sense
and antisense gene cassettes ....................................................................... 28
2.1 Introduction .............................................................................................. 29 2.2 Materials and Methods............................................................................. 33
2.2.1 Cloning Strategy ................................................................................ 33
2.2.2 Plasmids............................................................................................ 34
2.2.2.1 The ACC oxidase and ACC synthase cDNAs ............................. 34
2.2.2.2 Binary Vectors ............................................................................ 35
2.2.2.3 Plasmid extraction....................................................................... 35
5
2.2.3 Bacterial strains and media ............................................................... 38
2.2.4 Enzymes and DNA ............................................................................ 38
2.2.5 Cloning Strategy ................................................................................ 41
2.2.5.1 Removing gus from pBI121 ........................................................ 41
2.2.5.2 Replacing gus with synthetic linker ............................................. 41
2.2.5.3 Transferring CaMV 35S-linker-nos cassette from pBI121 to
pSCV1.0 ................................................................................................. 43
2.2.5.4 Inserting the Brassica oleracea ACC oxidase 1 and 2 cDNAs into
pJ1.0 by blunt-ended cloning .................................................................. 44
2.2.5.5 pMOSBlue kit .............................................................................. 46
2.2.5.6 Inserting Brassica oleracea ACC synthase cDNAs into pJ1.0 ... 46
2.2.5.7 Checking transformants for inserts ............................................. 47
2.2.5.8 Control Plasmid .......................................................................... 50
2.2.5.9 Producing and storing E. coli and A. rhizogenes stocks with the 7
constructs ............................................................................................... 51
2.3. Results .................................................................................................... 52 2.4. Discussion............................................................................................... 61
3.0 Transformation of GDDH33 with Agrobacterium rhizogenes ................... 63
3.1 Introduction .............................................................................................. 64
3.2 Materials and Methods............................................................................. 69
3.2.1 Plant material and culture conditions................................................. 69
3.2.2 Plant transformation .......................................................................... 69
3.2.3 Identification of transformed ‘hairy roots’ ........................................... 71
3.2.4 GUS detection ................................................................................... 71
3.2.5 PCR................................................................................................... 71
6
3.2.6 Shoot Regeneration from established root clones selected by GFP
fluorescence ............................................................................................... 74
3.2.7 DNA extraction .................................................................................. 74
3.2.8 Endonuclease restriction digest of genomic DNA.............................. 75
3.2.9 Southern Blot..................................................................................... 76
3.2.10 Hybridisation between N+ filters and 32P dCTP labelled DNA
probes ........................................................................................................ 77
3.2.11 Seeds .............................................................................................. 78
3.2.12 Statistical Analysis........................................................................... 78 3.3 Results ..................................................................................................... 79
3.3.1 Constructs ......................................................................................... 79
3.3.2 Inoculation treatment ......................................................................... 80
3.3.3 gfp reporter gene ............................................................................... 82
3.3.4 Co-transformation.............................................................................. 82
3.3.5 Regeneration ..................................................................................... 84
3.3.6 Determining stable insertion of the T-DNAs into the plant genome ... 86
3.3.7 Transgenic plant phentoype ............................................................ 102
3.3.8 Seed production of the transgenic lines .......................................... 105
3.4 Discussion.............................................................................................. 110
4.0 Analysis of broccoli bud post-harvest ethylene production and chlorophyll
levels............................................................................................................ 120
4.1 Introduction ............................................................................................ 121 4.2 Materials and Methods........................................................................... 127
4.2.1 Harvested broccoli heads ................................................................ 127
4.2.2 Ethylene Measurements.................................................................. 129
VII
4.2.3 Chlorophyll Measurements.............................................................. 130
4.2.4 Measuring dry weight of broccoli buds ............................................ 131
4.2.5 Data Analysis .................................................................................. 131 4.3 Results ................................................................................................... 133
4.3.1 Tissue to be assayed ...................................................................... 133
4.3.2 Controls for standardising the analysis of ethylene and chlorophyll 133
4.3.2.1 Calculating ethylene concentrations ......................................... 133
4.3.2.2 Analysis of pre/post harvest ethylene production and chlorophyll
levels in broccoli buds of non-transformed GDDH33 ............................ 134
4.3.3 Analysis of post-harvest ethylene production and chlorophyll levels of
non-transformed GDDH33 and the transgenic lines ................................. 136
4.4 Discussion.............................................................................................. 150 5.0 Identifying Brassica oleracea ACC oxidases ......................................... 159
5.1 Introduction ............................................................................................ 160
5.2 Materials and Methods........................................................................... 163
5.2.1 PCR................................................................................................. 163
5.2.2 Bacterial Artificial Chromosomes (BACs) ........................................ 165
5.2.3 Southern Blots and fliters ................................................................ 166
5.2.4 Hybridisation between N+ filters and
32p dCTP labelled DNA probes
................................................................................................................. 167
5.2.5 Genetic Mapping ............................................................................. 167
5.2.6 Cloning and sequencing PCR fragments ........................................ 168
5.2.7 Analyses of sequences ................................................................... 169 5.3 Results ................................................................................................... 170
8
5.3.1 Identifying Brassica oleracea BACs containing ACC oxidase genes
................................................................................................................. 170
5.3.1.1 Isolating BACs containing ACC oxidases ................................. 170
5.3.1.2 Identifying ACC oxidases from BACs isolated .......................... 171
5.3.2 Genetic Mapping of Brassica oleracea aco1 and aco2 genes ......... 180
5.3.3 Analysis of nucleotide sequences of closely related genes to Brassica
oleracea aco1 and aco2 ........................................................................... 186
5.4 Discussion.............................................................................................. 187 6.0 General Discussion ................................................................................ 190
6.1 Assessing the strategies ........................................................................ 191
6.1.1. Constructs ...................................................................................... 191
6.1.2 Transformation and regeneration .................................................... 193
6.1.3 Ethylene and chlorophyll ................................................................. 195
6.1.4 The aco1 and aco2 genes ............................................................... 196
6.2 A role for ethylene .................................................................................. 197
6.3 Future work ............................................................................................ 198
6.3.1 Fully understand the role and expression of the aco and acs genes
................................................................................................................. 198
6.3.2 Biochemical aspect ......................................................................... 198
6.3.3 Improved silencing of aco and acs in broccoli ................................. 199
6.3.4 Conventional breeding .................................................................... 200
6.3.5 Other ways to improve shelf-life ...................................................... 200
7.0 Appendices ............................................................................................ 202
8.0 References............................................................................................. 210
9
List of Figures Figure 1.1 The Agrobacterium infection process............................................. 8
Figure 1.2 Biosynthesis of ethylene .............................................................. 16
Figure 1.3 Proposed model of the ethylene signal transduction pathway ..... 17
Figure 1.4 The proposed mechanism for gene silencing by antisense mRNA
................................................................................................................ 20
Figure 2.1 Circular map of pBI121 ................................................................ 36 Figure 2.2 Circular map of pSCV1.0 ............................................................. 37
Figure 2.3 Schematic diagram of primer pairings for cDNA inserts in
antisense and sense orientations for aco1, aco2 and acs. ..................... 49
Figure 2.4 (a) T-DNA region of pJ1.0 with restriction sites. (b) Restriction
endonuclease digest confirming the integrity of the backbone plasmid,
pJ1.0 ....................................................................................................... 52
Figure 2.5 PCR products to determine presence and orientation of insert .... 54 Figure 2.6 PCR products to determine orientation of aco2 and acs inserts. 55
Figure 2.7 Confirmation of cDNA inserts ....................................................... 57
Figure 2.8 Confirmation of cDNA inserts and nos terminator ........................ 59
Figure 3.1 The TL-DNA of the agropine Ri plasmid pRi1855::GFP. .............. 67 Figure 3.2 The effect of treatment on construct transformation efficiency. .... 81
Figure 3.3 Nine-week-old GFP fluorescing root/callus. ................................. 83
Figure 3.4 Six-week-old GUS positive co-transformed root. ......................... 83
Figure 3.5 The morphogenesis from root callus to plantlet. .......................... 85
Figure 3.6 PCR of genomic DNA from the transgenic plants to confirm
insertion of the T-DNA constructs from the binary plasmid. .................... 88
10
Figure 3.7 PCR of the gfp gene from genomic DNA of the transgenic lines to
confirm insertion of the TL-DNA region of the virulence plasmid. ............ 88
Figure 3.8 Genomic DNA from aco1 and aco2 transgenic lines digested with
SstI restriction endonuclease and run on a 0.8% (w/v) agarose gel at 50V
for 18h. ................................................................................................... 93 Figure 3.9 Genomic DNA from acs and GUS transgenic lines digested with
SstI restriction endonuclease and run on a 0.8% (w/v) agarose gel at 50V
for 18h. ................................................................................................... 94
Figure 3.10 Autoradiograph of the hybridisation between the Brassica
oleracea L. var. italica aco1 283bp 3’UTR DNA probe and the digested
genomic DNA of Figure 3.8..................................................................... 95
Figure 3.11 Autoradiograph of the hybridisation between the Brassica
oleracea L. var. italica aco2 252bp 3’UTR DNA probe and the digested
genomic DNA of Figure 3.8..................................................................... 96
Figure 3.12 Autoradiograph of the hybridisation between the Brassica
oleracea L. var. italica acs full length cDNA probe and the digested
genomic DNA of Figure 3.9..................................................................... 97
Figure 3.13 Autoradiograph of the hybridisation between the 390bp gus DNA
probe and the digested genomic DNA of Figure 3.9. .............................. 98 Figure 3.14 Autoradiograph of the hybridisation between the 494bp GFP DNA
probe and the digested genomic DNA of Figure 3.8. .............................. 99
Figure 3.15 Autoradiograph of the hybridisation between the 494bp GFP DNA
probe and the digested genomic DNA of Figure 3.9. ............................ 100
Figure 3.16 Heading broccoli plants that have been regenerated from GFP
fluorescing roots, through a tissue culture phase.................................. 103
11
Figure 3.17 Determination of gfp inserts by PCR of 67 28/00 aco1A 3 T1
progeny grown from seed. .................................................................... 107
Figure 3.18 Determination of aco1A inserts by PCR of 67 28/00 aco1A 3 T1
progeny grown from seed. .................................................................... 108
Figure 4.1 (a) Gas chromatograph; (b) Glass container with broccoli buds
evolving ethylene; (c) Injecting 1ml air sample taken from (b); (d)
Chlorophyll diffusing from broccoli buds in 5mls methanol. .................. 132
Figure 4.2 The rate of diffusion of chlorophyll at 20oC in the dark, from 0.2g
broccoli buds in 5ml methanol. ............................................................. 134 Figure 4.3 Ethylene production by harvested/non-harvested untransformed
GDDH33 broccoli buds. ........................................................................ 135
Figure 4.4 Chlorophyll levels of harvested/non-harvested untransformed
GDDH33 broccoli buds. ........................................................................ 136
Figure 4.5 Post-harvest ethylene production and chlorophyll levels of buds
from untransformed GDDH33. .............................................................. 144
Figure 4.6 Post-harvest ethylene production and chlorophyll levels of buds
from GDDH33 transformed with gus control constructs. ....................... 144
Figure 4.7 Post-harvest ethylene production and chlorophyll levels of buds
from GDDH33 transformed with ACC oxidase 1 antisense constructs. 145
Figure 4.8 Post-harvest ethylene production and chlorophyll levels of buds
from GDDH33 transformed with ACC oxidase 1 sense constructs. ...... 145
Figure 4.9 Post-harvest ethylene production and chlorophyll levels of buds
from GDDH33 transformed with ACC oxidase 2 antisense constructs. 146
Figure 4.10 Post-harvest ethylene production and chlorophyll levels of buds
from GDDH33 transformed with ACC oxidase 2 sense constructs. ...... 146
XII
Figure 4.11 Post-harvest ethylene production and chlorophyll levels of buds
from GDDH33 transformed with ACC synthase antisense constructs. . 147
Figure 4.12 Post-harvest ethylene production of buds from two lines of
GDDH33 transformed with ACC synthase antisense constructs. ......... 147 Figure 4.13 Post-harvest ethylene production and chlorophyll levels of buds
from GDDH33 transformed with ACC synthase sense construct.......... 148
Figure 4.14 Relationship between post-harvest ethylene production and
chlorophyll loss of buds from transformed and untransformed GDDH33.
.............................................................................................................. 148 Figure 4.15 Broccoli heads post- harvest. ................................................... 149
Figure 5.1 Schematic diagram of the predicted Brassica oleracea L. var.
italica ACC oxidase 1 gene................................................................... 165
Figure 5.2 Schematic diagram of the predicted Brassica oleracea L. var.
italica ACC oxidase 2 gene................................................................... 165
Figure 5.3 Autoradiograph showing hybridisation between the Brassica
oleracea BAC library filter 1, (a) and filter 2, (b) and the Brassica oleracea
var. italica ACC oxidase 2 full length cDNA (Pogson et al., 1995a). ..... 170
Figure 5.4 Autoradiograph showing hybridisation between the B. oleracea
ACC oxidase 1 cDNA (Pogson et al., 1995a) 3’ UTR 283bp probe and the
EcoRI RFLP BAC filter.......................................................................... 172
Figure 5.5 Autoradiograph hybridisation between the B. oleracea ACC
oxidase 2 cDNA (Pogson et al., 1995a) 3’ UTR 252bp probe and the
EcoRI RFLP BAC filter.......................................................................... 173
13
Figure 5.6 Electrophoretogram showing the products of a PCR between the
BAC DNA and aco1 specific primers (2 and 3) and aco2 specific primers
(2 and 4). .............................................................................................. 175
Figure 5.7 Brassica oleracea ACC oxidases 1 and 2 intron 1 aligned to
determine possible homology. .............................................................. 176
Figure 5.8 Brassica oleracea ACC oxidases 1 and 2 intron 2 aligned to
determine possible homology. .............................................................. 176
Figure 5.9 Electrophoretogram showing the PCR products from primers 5 and
7 on BAC’s 3, 4, 12 and 20. .................................................................. 177 Figure 5.10 Electrophoretogram showing PCR products from all BACs with
primers 5 and 7, and, 9 and 10. ............................................................ 178
Figure 5.11 Autoradiograph showing the hybridisation between the Brassica
oleracea BAC EcoRI RFLP filter and the Brassica oleracea aco1 intron 2
probe. ................................................................................................... 179
Figure 5.12 Autoradiograph showing the hybridisation between the B.
oleracea ACC oxidase1 cDNA 3’ UTR 283bp probe and the AG mapping
filter. ...................................................................................................... 181
Figure 5.13 Autoradiograph showing the hybridisation between the B.
oleracea ACC oxidase1 cDNA 3’ UTR 283bp probe and the NG mapping
filter. ...................................................................................................... 182
Figure 5.14 Autoradiograph showing the hybridisation between the B.
oleracea ACC oxidase2 cDNA 3’ UTR 252bp probe and the AG mapping
filter. ...................................................................................................... 183
14
Figure 5.15 Autoradiograph showing the hybridisation between the B.
oleracea ACC oxidase2 cDNA 3’ UTR 252bp probe and the NG mapping
filter. ...................................................................................................... 184
Figure 5.16 The linkage map positions of aco1 and aco2 in Brassica oleracea
for the loci that could be scored by EcoRI RFLP polymorphism. .......... 185
15
List of tables Table 1.1 Genetically modified traits on the horizon ........................................ 5
Table 2.1 Oligonucleotide sequences used in these experiments ................. 39
Table 2.2 PCR reaction components............................................................. 40
Table 2.3 PCR program................................................................................. 40
Table 2.4 Predicted PCR product sizes for the insertion of ACC cDNAs in
sense or antisense orientations into pJ1.0.............................................. 48
Table 3.1 Oligonucleotide sequences used as primers for the verification of T-
DNA insertion and making DNA probes for hybridisation. ...................... 72
Table 3.2 PCR reaction components............................................................. 73 Table 3.3 PCR program................................................................................. 73
Table 3.4 Transformation efficiency of LBA 9402 pRi1855::GFP carrying
binary plasmids with the B. oleracea var. italica ACC cDNA constructs on
GDDH33. ................................................................................................ 79
Table 3.5 Transgenic lines regenerated into whole plants............................. 84 Table 3.6 Summary of the PCR and autoradiograph data for confirming the
presence of T-DNAs in lines regenerated through tissue culture.......... 101
Table 3.7 The effect of the rol genes on transgenic plant morphology. ....... 104 Table 3.8 The total numbers of transgenic plants produced from all constructs
with the effect of the rol genes on mortality. ......................................... 105
Table 3.9 The number of seeds recovered from transgenic lines with their
particular rol phenotype ........................................................................ 106
Table 3.10 Segregation of T-DNA inserts in the T(1) progeny of 28/00 aco1A 3
.............................................................................................................. 109
16
Table 4.1 Transgenic lines regenerated by tissue culture ........................... 128 Table 4.2 The concentration of chlorophyll (Chl) in the cuvette was calculated
using simultaneous equations, based on the absorbance readings at 645
and 663nm (Hipkins and Baker, 1986). ................................................ 131
Table 4.3 Ethylene levels of GDDH33 buds post-harvest as nanolitres per
gram bud (dry weight) per hour............................................................. 137
Table 4.4 Chlorophyll values of GDDH33 buds post-harvest as micrograms
per gram bud (fresh weight).................................................................. 137
Table 5.1 Oligonucleotide sequences used for primers to determine gene
polymorphisms and produce DNA probes . .......................................... 163
Table 5.2 PCR reaction components........................................................... 164 Table 5.3 PCR program............................................................................... 164
Table 5.4 The codes of the B. oleracea BAC colonies that hybridised to the B.
oleracea ACC oxidase 2 full length cDNA probe. ................................. 171
1
1.0 General Introduction
2
1.1 Introduction
Brassica oleracea is a highly polymorphic species encompassing a
wide range of important vegetable fodder crops, including cabbage (var.
capitata), brussels sprouts (var. gemmifera), broccoli (var. italica), cauliflower
(var. botrytis), kohlrabi (var. gongylodes), and kale (vars, medullosa, ramosa
and acephala) (Hodgkin, 1995). It is a diploid species with 2n = 2x =18, and a
nuclear DNA content of 1.81 x 10-12g (Verma and Rees, 1974). Brassica
oleracea is an outbreeding species with a sporophytic self-incompatibility
system (Thompson, 1957). The wild species is commonly biennial or
perennial, but annual crops also exist such as summer cauliflower and
calabrese (Hodgkin, 1995).
The general view has long been that the early evolution of the different
cultivated types took place in the mediterranean area (Helm, 1963). However,
while much of the early selection may have occurred here, the evidence
currently suggests that the species from which modern crops are derived is
the wild B. oleracea and not the mediterranean species. As a result of their
molecular studies, Song et al. (1988) propose that the cultivated morphotypes
originated from a single ancient progenitor that was similar to wild B. oleracea.
This then became widely spread along the coasts of the Mediterranean and
North Atlantic and the different forms evolved in different areas through
selection and adaptation to various climates (Hodgkin, 1995).
The term ‘broccoli’ has been applied to the young floral shoots of
several Brassica crops as diverse as cabbages and turnips, but usually refers
to heading and sprouting forms belonging to the Brassica oleracea L. var.
italica group. These include Purple, and White Sprouting Broccoli, Purple
3
Cape, Purple Sicilian, and Black Broccoli varieties (Gray, 1993). The
calabrese-broccoli is now the major crop within the italica group and in
breeding terms, the most developed (Gray, 1982). It is derived from biennial,
Italian green sprouting broccoli from the Calabrian region of Italy (Crisp et al.,
1986). Calabrese has become a specialist crop in the USA, Japan and
Europe, and is marketed as a quality vegetable (Gray, 1993). There now exist
many cultivars, the majority of which are F1 hybrids. These have been
intensively selected for rapid maturity so that the crop is now effectively an
annual one with an increased terminal head size, which has changed the crop
from a sprouting plant virtually to a heading one. Gray (1993) states that
improvements in quality and extending the maturity range of horticultural
crops will be more important than further increases in productivity.
Broccoli is one of the most successful ‘new’ crops in the United
Kingdom. The crop value of home produced marketed broccoli has doubled
from £18.9m to £44.9m in the last 10 years (DEFRA statistics, 2000), where
there has been an overall decline of marketed Brassicas. The home grown
crop does not satisfy demand and there are significant imports (29,000 tonnes
with a value of £14m). The UK crop is either processed (frozen) or marketed
fresh. There is convincing information from industry that crops of the same
variety treated in the same way after harvest, vary in their post-harvest
performance (D. Pink, pers. comm. HRI-Wellesbourne). This lack of
predictability results in the major retailers specifying a product life of only two
days from delivery to ‘purchase before date’, and a further two days ‘best
consumed by’ in an attempt to ensure a ‘quality’ product for the consumer.
4
This short post-harvest life results in significant waste for the retailer (a cost
borne by the consumer) and consumers themselves.
1.2 Genetic Improvement of Crop Plants
Genetic improvement of crop species has always been a major goal in
breeding programmes. Sexual crossing between species and selection of
desirable traits has led to the high yielding varieties of the present day.
However, this may take many generations and time to remove unwanted
characteristics and is limited by inter-species barriers. Genetic manipulation
by transformation has been developed over the past twenty years as a means
to facilitate direct gene transfer between two unrelated species as well as
within species. Dominant useful agronomic traits may be transferred directly
and rapidly offering potential for both gene function studies and crop
improvement strategies. Table 1.1 lists the main areas targeted for genetic
improvement by direct gene transfer.
5
Table 1.1 Genetically modified traits on the horizon (after Chrispeels and Sadava, 1994)
Pest and Weed Management
Herbicide tolerance virus resistance Insect resistance Bacterial resistance Fungal resistance
Plant Breeding
Male sterility High-methionine and high-lysine seeds Forage rich in sulphur amino acids
Agronomic Properties
Altering cold sensitivity Improving salt tolerance Improving water stress tolerance
Molecular Farming
Oils Starch Plastic Enzymes Pharmaceuticals
Postharvest Qualities
Delay of fruit ripening Delay of flower senescence High-solids tomatoes High-starch potatoes Sweeter vegetables
Detoxifying contaminated soils
At present, in the United States there have been environmental field
releases for Brassica oleracea crops with insect resistance and male sterility
genes. The insect resistance gene (CryIA) is derived from the soil bacterium
Bacillus thuringiensis (Bt). It codes for a crystal endotoxin which when
ingested by the target pest causes disruption of the gut wall leading to
eventual starvation (Aronson and Shai, 2001). The advantage of inserting the
gene into the crop is that it targets the pest specifically unlike pesticides, and
does not affect other species of insects. However, like chemical pesticides it
also provides a constant selection pressure for the pest species to overcome
and will therefore need to be managed in an integrated system (Weisz et al.,
1994). Introduction of male sterility genes into Brassica oleracea is useful for
6
the production of hybrid seeds that may produce higher yielding plants. It is
difficult to produce hybrid seeds in Brassica because the male and female
strucures are present in the same flower. American seed companies have
used an RNase gene from Bacillus amyloliquefaciens (Barnase) to destroy
the RNA in the pollen making cells, and thus making the plants male sterile
(http://www.nbiap.vt.edu/cfdocs/fieldtests3.cfm).
At present, it is estimated that there are hundreds of transgenic
varieties commercially available, mostly of soybean, corn, cotton and canola.
According to the International Service for the Acquisition of Agri-biotech
Applications (ISAAA), the global area of transgenic crops increased by more
than 25-fold, from 1.7 million hectares in 1996 to 44.2 million hectares in
2000. Of the top four countries that grew 99% of the global transgenic crop
area, the USA grew 68%, Argentina 23%, Canada 7%, and China 1%
(James, 2000). In 2000, over 99% of the transgenic acreage was devoted to
just two traits – herbicide and insect resistance.
1.3 Transformation of Brassica
Brassica and related species are fairly easy to handle in tissue culture
systems making this genus a suitable subject for exploiting techniques for
genetic transformation for breeding purposes (Poulsen, 1996). The first step
is to introduce ‘foreign’ DNA into the cells/tissue of interest. This is followed by
selection of the transformed cells on a suitable medium with plant growth
regulators to induce regeneration into a whole novel plant.
Techniques developed to introduce DNA into Brassica plant cells have
involved direct gene transfer, in which naked DNA is introduced into
7
protoplasts or intact cells via microinjection or electroporation; bombardment
with DNA coated particles using a particle gun (Puddephat et al., 1996); or to
utilise the natural gene transfer capacity of Agrobacterium. The latter has
been the most effective method for producing stable transformants that
segregate in Mendelian fashion.
Crown gall disease and hairy root disease are neoplastic growths on
plants incited by virulent strains of Agrobacterium tumefaciens and
Agrobacterium rhizogenes, respectively. Transfer DNA or (T-DNA) is carried
on the tumour-inducing (Ti) plasmid of A. tumefaciens and root-inducing (Ri)
plasmid of A. rhizogenes. The T-DNA is transferred to the plant where it
integrates into the plant nuclear DNA (Gelvin, 1990).
Integration of the T-DNA into the host genome (Figure 1.1) depends on
a complex series of chemical signals between the Agrobacterium and the
plant (Gelvin, 2000). These signals include neutral and acidic sugars, phenolic
compounds, opines, Vir (virulence) proteins, and the T-DNA that is ultimately
transferred from the bacterium to the plant cell (Gelvin, 2000). The T-DNA
transfer process initiates when Agrobacterium perceives certain phenolic and
sugar compounds from wounded cells (Hooykaas and Beijersbergen, 1994).
In the presence of these phenolic compounds, the VirA sensory protein
becomes phosphorylated. The active VirA protein then transphosphorylates
the VirG protein which initiates vir gene transcription (Jin et al., 1990a,b).
Most of the induced Vir proteins are directly involved in T-DNA processing
from the Ti/Ri- plasmids, and the subsequent transfer of T-DNA from the
bacterium to the plant (Gelvin, 2000). VirD1 and VirD2 are responsible for
nicking the Ti/Ri-plasmids at 25bp directly repeated sequences called T-DNA
8
borders that flank the T-DNA (Filichkin and Gelvin, 1993). The VirD2 protein
strongly associates with the nicked strand, so that the T-DNA enters the plant
cell as protein/nucleic acid complex (Gelvin, 2000). It is probable that the
VirD2 and VirE2 protein are important in guiding the T-strand into the plant
cell nucleus and integrating it into the genome (Gelvin, 2000).
Figure 1.1 The Agrobacterium infection process. Critical steps that occur to or within the bacterium (chemical signalling, vir gene induction, and T-DNA processing) and within the plant cell (bacteria attachment, T-DNA transfer, nuclear targeting, and T-DNA integration) are highlighted along with genes and/or proteins known to mediate these events (after Gelvin, 2000).
It is advantageous to distingiush phenotypically similar transformed
plants from non-transformed plants by means of selection. Selection requires
9
the use of genes that function as selectable markers to permit the recovery of
transformed cells and tissues, and genes that act as reporters of transgene
expression. The most common selectable markers are based on antibiotic
resistance genes, such as nptII for neomycin phosphotransferase, which
detoxifies a number of aminoglycoside antibiotics, including neomycin and
kanamycin (Puddephat et al., 1996). The most widely used reporter gene, the
gus gene encoding for -glucuronidase, allows in situ histological assessment
of transgene expression, but the assay is toxic to the plant. A potentially
promising in vivo reporter gene is the green fluorescent protein (GFP) from
the bioluminescent jellyfish Aequorea victoria (Sheen et al., 1995). GFP
expression is cell autonomous and independent of cell type and location.
Fluorescence is achieved by excitation of the GFP protein with blue or UV
light, emitting visible light that may be observed by the naked eye. GFP does
not require any exogenous substrates (Chrispeels and Sadava, 1994;
Delagrave et al., 1995; Gray et al., 1992, Heim et al., 1994) or produce an
endotoxic product when targeted to the endoplasmic reticulum (Haseloff et al.,
1997). Therefore, histological in vivo studies may be carried out without loss
of tissue.
The choice of regulatory sequences for the control of transgene
expression is still somewhat limited. Most experiments on engineered pest
and disease resistance have used constitutive promoters, most often the
cauliflower mosaic virus (CaMV) 35S sequence. Ideally, plant-derived
promoters may be more acceptable to regulatory authorites than virus derived
promoters (Pierpoint, 1995), but would need to offer the same advantages.
10
The most successful Brassica tissue for transformation and
regeneration appears to be that of explants composed of, or derived from,
meristematic tissues at the early stages of plant development (Puddephat et
al., 1996). Cells transformed by wild-type A. rhizogenes strains may be readily
identifiable by virtue of hairy root formation. These roots are cellular clones
originating from a single transformed cell (Tempe and Casse-Delbart, 1987)
and are capable of regenerating into whole plants (Tepfer, 1990). In most
cases, B. oleracea explants inoculated with A. rhizogenes form roots between
one and three weeks after inoculation (Hosoki et al., 1991). Typically, hairy
roots exhibit rapid, phytohormone independent growth, with extensive
plagiotropic branching. The regeneration of shoots from hairy-roots is
achieved by the treatment of cytokinins and auxins. Regeneration frequencies
of root clones from broccoli are low, averaging 11% (Hosoki et al., 1991) (for a
full review see Puddephat et al., 1996).
The Agrobacterium rhizogenes Ri-plasmid is necessary for the
successful transformation of the host plant. Outside the TL(left) -DNA borders
the Ri-plasmid carries genes specific for TL -DNA transfer, such as the Vir
genes and, within the borders, genes that elicit neoplastic root growth. The TL
region includes four genes: rolA, rolB, rolC, rolD (see Figure 3.1)(White et al.,
1985; Zambryski et al., 1989). The first three are important for hairy root
production from transformed cells, and the morphological aberrations
observed in transformed plants (Schmulling et al., 1988). In most plants rolB
is capable of inducing root formation on its own (Schmulling et al., 1988).
Although rolA and rolB can induce aberrations in transformed plants, it
appears that rolC is responsible for many of the most severe features
11
including reduced chlorophyll content, smaller than normal wrinkled leaves,
dwarfing, loss of apical dominance and various forms of flowering
abnormalities (Schmulling et al., 1988). The deleterious effects of rolC may be
avoided by developing an A. rhizogenes vector system in which this gene has
been eliminated. Senior et al., (1995) used a rolA mutant to transform
Antirrhinum, producing a semi-dwarf phenotype. This approach with a rolC
mutant may lead to the production of phenotypically normal transgenic plants.
As a scientifically and more socially acceptable approach, co-
transformation in which marker genes and genes of interest are located on
separate T-DNAs, allows the use of a selectable marker during plant
regeneration followed by recovery of progeny which contain the desired
gene(s) but lack a marker gene (Puddephat et al., 2001). If transgenes have
been inserted into the plant genome at physically unlinked loci, the gene of
interest can be segregated from the selectable marker in the next generation.
Independent T-DNA insertion has been achieved with co-transformation,
using one plasmid with multiple T-DNAs (Depicker et al., 1985; Komari et al.,
1996) or separate plasmids with different T-DNAs that are contained in either
one (Komari et al., 1996) or more Agrobacterium strains (Depicker et al.,
1985). Puddephat et al. (2001) used a co-transformation strategy based on
one Agrobacterium strain, harbouring two plasmids, each carrying separate T-
DNAs. Puddephat et al. (2001) recovered phenotypically normal transgenic
Brassica oleracea plants from Agrobacterium rhizogenes-mediated co-
transformation and selection of transformed hairy roots by the GUS assay.
There had been sufficient independent integration of T-DNAs at unlinked sites
12
for the marker genes to segregate from the genes of interest in the progeny of
the T0 plants.
1.4 Senescence of Broccoli
Senescence is a complex, highly regulated plant developmental phase
that results in the co-ordinated degradation of molecules. Genes known to be
upregulated in this process encode degradative enzymes such as proteases
and nucleases, and enzymes involved in lipid and carbohydrate metabolism.
This leads to an ordered series of events involving cessation of
photosynthesis, breakdown of chlorophyll and removal of amino acids
(Buchanan-Wollaston, 1997).
Broccoli is a floral vegetable that is harvested when the flowering
heads are immature and growing rapidly. The head of broccoli is composed of
hundreds of immature florets arranged in whorls on a fleshy stem. Each
broccoli floret is composed of male and female reproductive structures
surrounded by immature petals and enclosed within chlorophyll-containing
sepals (Pogson et al., 1995a). At the time of harvest the florets are closed and
unpollinated. Vegetables harvested when immature and before growth has
ceased experience a disruption of energy, nutrient, and hormone supplies and
consequently senesce rapidly (King and Morris, 1994a,b). An estimate by
Pogson and Morris (1997) is that 1% of dry mass is respired per hour after
harvest. Fresh broccoli is therefore highly perishable, with a storage life of 3 to
4 weeks in air at 0oC (Makhlouf et al., 1989) and 2 to 3 days in air at 20oC (Wang, 1977). As broccoli deteriorates, the head yellows due to degradation
of chlorophyll (Wang, 1977). The characteristic sepal yellowing is the first
13
visual sign of senescence but the final stage of this process. The major
catabolic breakdown and oxidation of macromolecules has already occurred
leaving the tissue with poor nutrient status, recognised by the consumer. The
branchlets and florets also lose turgor and become flaccid due to water loss
(Brennan and Shewfelt, 1989). King and Morris (1994b) found that the stress
imposed by harvest, trimming, transportation and washing procedures greatly
accelerated senescence.
Lieberman and Hardenburg (1954) suggested that yellowing in broccoli
florets may be due to endogenous ethylene. There was no visible yellowing of
broccoli placed in nitrogen atmospheres, under which conditions ethylene was
not evolved. The proposition by Lieberman and Hardenburg was supported by
Wang (1977) who delayed chlorophyll loss using ethylene biosynthetic
inhibitors (aminoethoxyvinylglycine -AVG) and inhibitors of ethylene action
(silver ions). Tian et al. (1994) observed that removing the reproductive
structures (stamens and pistil) greatly reduced sepal yellowing. The
reproductive structures produced more than double the amount of ethylene
than other tissues in the floret.
1.5 Ethylene Biosynthesis
Ethylene is produced by plant tissues in amounts ranging from almost
none up to 500 nl/g/h (Burg, 1962). It is biologically active in trace amounts
(as little as 10-100 nl/l of air). Ethylene production is induced during several
developmental stages, including fruit ripening, seed germination, leaf and
flower senescence, and abscission. It is also induced by external factors, such
as wounding, anaerobiosis, viral infection, auxin treatment, chilling injury,
14
drought, Cd2+ and Li+ ions and mechanical stress (Abeles, 1973; Yang and
Hoffman, 1984). Ethylene biosynthesis occurs via the enzymes ACC
synthase and ACC oxidase, which catalyse the conversions of S-
adenosylmethionine to 1-aminocyclopropane-1-carboxylic acid (ACC) and
ACC to ethylene, respectively (see Figure 1.2) (Yang and Hoffman, 1984).
1.5.1 ACC synthase
After the discovery that ACC is the immediate precursor of ethylene, it
became clear that ACC synthase played an important role in regulating the
production of ethylene. The induction of ethylene production by a variety of
means was reported to be due to de novo synthesis of this enzyme (Kende,
1989), although other reports state that its activity may be post-translationally
controlled (Morgan and Drew, 1997). There is an emerging picture that ACC
synthase is encoded by a highly divergent multigene family. In tomato, for
example, ACC synthase is encoded by at least seven genes, two of which are
expressed during fruit ripening (Van der Straeten et al., 1990; Morgan and
Drew, 1997; Oetiker et al., 1997). Similarly, in Arabidopsis and rice, ACC
synthase genes are differentially expressed in response to developmental,
hormonal and environmental stimuli (Liang et al., 1992).
1.5.2 ACC oxidase
ACC oxidase is constitutively expressed in most vegetative tissues
(Yang and Hoffman, 1984) and is induced to higher levels during fruit ripening
(Gray et al., 1992). In senescing broccoli, the concomitant increase of ACC
oxidase activity was suppressed by cycloheximide (an inhibitor of translation),
15
suggesting increased ACC oxidase activity results from the de novo synthesis
of the enzyme protein, as with ACC synthase (Kasai et al., 1998). A cDNA
encoding ACC oxidase was first cloned from tomato fruit (Slater et al., 1985).
It encodes a 37KDa dioxygenase that belongs to the superfamily of
Fe2+/ascorbate oxidases, which also require CO2 and O2 as co-substrates
(McGarvey et al., 1992). 1.6 Signalling
Molecular genetic analysis of ethylene signalling in Arabidopsis
thaliana, has led to the identification of a number of genetic loci that are
required for normal ethylene responses (McGrath and Ecker, 1998). The
elucidation of this biochemical pathway has been achieved with ethylene
insensitive mutants and mutants with constitutive ethylene responses. A
mutation in the ETR1 gene allowed the elucidation that ethylene is perceived
by a histidine kinase membrane bound receptor protein (see Figure 1.3).
Ethylene binding results in the inactivation of a MAP kinase cascade that is
regulated by another kinase, CTR1. Inactivation of this inhibitory MAP kinase
cascade allows activation of the ethylene signalling cascade. It is proposed
that this cascade amplifies the ethylene signal and leads to the transcription of
target genes that play important roles in developmental phases including
senescence.
16
Figure 1.2 Biosynthesis of ethylene (after Yang and Hoffman, 1984)
17
ETR1
C2H4
Phosphory -lates MEK
CTR CTR1
Phosphory- lates MAPK
Inhibits cascade without ethylene
MEK
MAPK
EIN2
MEK
MAPK
EIN2
EIN5
EIN6
EIN3, EIL1, EIL2, EIL3(?)
Transcription factors
Target Genes
Figure 1.3 Proposed model of the ethylene signal transduction pathway (after McGrath and Ecker, 1998). Binding of ethylene by the receptor complex inhibits activity of the MAP kinase cascade, allowing transmission of the ethylene signal. CTR1=constitutive triple response; EIN = ethylene insensitive mutant; EIL = EIN-like; ETR= ethylene resistant; MAPK = mitogen avtivated protein kinase; MEK= MAPK kinase kinase
18
Using the Arabidopsis ethylene receptor ETR1 as a probe, Payton and
co-workers (Payton et al., 1996) isolated the tomato homolog tETR. They
showed that mRNA of etr1 was undetectable in unripe fruit or pre-senescent
flowers, but increased in abundance during the early stages of ripening, flower
senescence in abscission zones, and was greatly reduced in fruit of ripening
mutants deficient in ethylene synthesis or response. More recent work by
Sato-Nara et al. (1999) on muskmelon (Cucumis melo) also found a stage-
and tissue-specific expression of the ethylene receptor during fruit
development. They suggest that as the receptors have an inhibitory effect on
the signalling cascade, if more are expressed at a specific time there would
need to be more ethylene produced to compensate. It follows that if there was
a large ethylene burst, with little receptor expression, the cascade would be
activated, but possibly to a lesser extent than if there were greater numbers of
receptors. Also, if great numbers of receptors were expressed, but with little
ethylene, there would be potentially greater inhibition of the signal cascade.
Large numbers of ethylene receptor proteins are prevalent at the
climacteric point of fruit development in the melon fruit. There are other
reports indicating that receptors increase while tissue sensitivity of ethylene is
increasing (Vriezen et al., 1997; Wilkinson et al., 1995). In this light, it is
interesting that Makhlouf et al. (1989) described the post-harvest senescence
of broccoli as climacteric since florets became yellow as respiration and
ethylene production increased.
The work over the last 20 years with the elucidation of ethylene
biosynthesis and signalling pathway suggests that there are three main
controls for the physiological responses invoked by ethylene. These are ACC
19
synthase, known as the rate limiting enzyme in the biosynthetic pathway; ACC
oxidase, which catalyses the final substrate ACC into ethylene and; the
ethylene receptor which may have varying degrees of sensitivity to ethylene
depending on developmental phase of the plant.
1.7 Antisense Technology
Antisense technology has proved a powerful tool for understanding
genetic control in plants through gene silencing. The technique was
developed by D. Grierson and colleagues at Sutton Bonington (University of
Nottingham) to understand the genetics behind fruit ripening of tomato
(Lycopersicon esculentum). Hamilton et al. (1990) used a tomato ACC
oxidase cDNA known as pTOM13, inserted into a binary vector in the
antisense orientation, between the CaMV 35S promoter and its terminator.
Figure 1.4 illustrates the means by which antisense are thought to down-
regulate genes expression at the mRNA level.
20
Promoter
Promoter
(i) Sense gene
5’
3’
(ii) Sense mRNA
ATGTGCAAA TACACGTTT
AATAA TTATT
Termination
5’ AUGUGCAAA AAUAA 3’
(iii) Antisense gene: the coding region of the normal gene in reverse orientation
5’ TTTGCACAT 3’ AAACGTGTA
AATAA
TTATT
(iv) Antisense mRNA
5’ UUUGCACAU AAUAA 3’
(v) Double stranded mRNA formed between sense and antisense mRNA
5’ AUGUGCAAA AAUAA 3’
3’ AAUAA
UACACGUUU 5’
Figure 1.4 The proposed mechanism for gene silencing by antisense mRNA (after Bryant, 1988).
The theory illustrated in Figure 1.4 proposes that an RNA duplex forms
between the sense and antisense strands. The double stranded RNA
molecules are not translated (through the ribosomes) and are exposed to
native dsRNases within the cytoplasm. The silencing relies on an excess of
the antisense mRNAs in the cell to prevent translation of the normal mRNAs.
Hamilton et al. (1990) found that neither sense nor antisense transcripts were
detected in ripening fruit with the pTOM13 construct. ACC oxidase activity
was reduced by 93% in transgenic plants with two copies of the antisense
construct. This permitted normal development of fruit ripening without the
extreme softening, cracking and spoilage associated with over-ripening. In
21
addition, the senescence of the leaves was delayed for 7-10 days. Olson et al.
(1991) used a tomato ACC synthase antisense construct LE-ACC2.
Interestingly, this resulted in an almost complete inhibition of mRNA
expression for the two ACC synthase genes (LE-ACC2 and LE-ACC4), and
the fruits never ripened. The antisense phenotype can be reversed by
treatment with ethylene or propylene, an ethylene analog (Oeller et al., 1991).
The antisense approach has been used to down-regulate the ripening genes
responsible for the cell wall degrading enzymes polygalacturonase and
pectinesterase (Picton et al., 1995) leading to the first biotechnology product
based on gene silencing, the ‘Flavr Savr’ tomato, marketed in the United
States by Calgene in 1994 (Grierson et al., 1996).
1.8 Isolation of ACC genes
The initial experiment by Hamilton et al. (1990) with pTOM13 where an
antisense ACC oxidase gene inhibited the synthesis of the hormone ethylene
was a major step towards cloning the genes and characterising the elusive
ethylene biosynthetic enzymes. The original cloned DNA sequence has been
used as a cDNA probe to isolate homologs from numerous other species. The
major targets for gene isolation appear to be climacteric fruits such as
banana, melon, apple, pear and apricot (LopezGomez et al., 1997; Guis et al.,
1997; Bolitho et al., 1997; Kato and Hyodo, 1999; MbeguieAMbeguie et al., 1999). It is clear from a sequence analysis database (BLAST search,
http://www.ncbi.nlh.gov/BLAST/) that of these genes occur in a great number
of other species including Arabidopsis and tobacco, two genetic model plants
22
as well as crop species such as cereals (rice- Oryza sativa), legumes (bean- Phaseolus vulgaris) as well as Brassica species.
Following the isolation of these genes there have been a number of
antisense gene-silencing experiments in tomato melon and tobacco to
understand gene function (Ayub et al., 1996; Knoester et al., 1997).
In 1995, B. Pogson, and co-workers C. Downs and K. Davies at the
Institute for Crop & Food Research, New Zealand isolated two ACC oxidase
genes and a cDNA clone encoding ACC synthase from Brassica oleracea
(Pogson et al., 1995a). The ACC synthase had a transcript of approximately
1700 bp with an open reading frame of 491 amino acids. The transcript
abundance did not change in florets post-harvest and was detected in similar
abundance in green and senescing tissues (Pogson et al., 1995b). Reports
suggest that this enzyme may be post-translationally regulated (Morgan and
Drew, 1997). It appears that an ACC synthase-inactivase exists which must
be phosphorylated for activity. This raises the possibility that some of the very
rapid effects on ethylene biosynthesis may initially occur by alteration of an
existing pool of ACC synthase.
Pogson et al. (1995a) used a cDNA clone encoding ACC oxidase from
Brassica juncea (mustard, Pua et al., 1992) to probe a broccoli (B. oleracea L.
cv. Shogun) 48h post-harvest cDNA library. Two cDNAs with high homology
to ACC oxidases in the databases were isolated and named aco1 and aco2.
The lengths of the aco1 and aco2 were 1237 and 1232bp, respectively.
Commencing at the first Met, the predicted open reading frames of aco1 and
aco2 were 320 and 321 amino acids, respectively. aco1 and aco2 shared
nucleotide identity of 83% in their putative translated regions and 48% in the
23
3’ untranslated regions. The short 5’ untranslated regions shared no homology.
1.9 Tissue specific expression of ACC oxidases in broccoli
Transcripts corresponding to aco1, but not aco2, were detected in
florets and, at low levels, in leaf tissues frozen immediately after harvest
(Pogson et al., 1995a). The abundance of both aco1 and aco2 transcripts
increased in florets during the post-harvest period and was accompanied by
an increase in ethylene production.
In stamens, neither aco1 nor aco2 transcripts were detected at harvest
(Pogson et al., 1995a). Transcripts for aco1 were also not detected in the
stamens post-harvest, whereas aco2 transcript levels increased markedly at
24 h and again at 48 h after harvest. In female reproductive structures, neither
transcript was detected at harvest. However, high levels of aco2 transcript
were detected at 24 h but not 48 h after harvest. A closer inspection showed
that aco2 transcripts significantly increased 2 h after harvest in whole florets
(Pogson et al., 1995a). Pogson et al. (1995a) results were supported by
evidence from Kasai et al. (1996) that ACC oxidase activity in florets rapidly
increased to reach a peak, followed by a sharp decline to a low level as
senescence progressed. This paralleled the pattern of ethylene production.
Pogson et al. (1995a) found a large increase in aco2 transcripts following ABA
treatment, suggesting that changes in water relations and ABA concentrations
may form part of the response to harvest.
These results indicated to Pogson et al. (1995a) that aco1 was unlikely
to initiate chlorophyll loss associated with broccoli senescence and that its
24
probable role was the generation of the ‘basal ethylene’ produced before
harvest. They suggested that the ethylene produced was a consequence of
rapid and large increases of aco2 transcripts within the reproductive
structures and this is one of the main signals influencing sepal yellowing.
In mature carnation, different parts of the flower produce disparate
amounts of ethylene during senescence (Woodson, 1987). In unpollinated
flowers, ethylene synthesis may occur simultaneously in all floral organs. After
pollination, however, there is a rapid induction of ethylene synthesis in the
style, followed a few hours later by greater ethylene synthesis in the petals,
leading to premature senescence (Nichols, 1977). The pattern of senescence
in carnation is similar to that of broccoli. It has been reported that reproductive
development is often associated with leaf senescence in intact plants (Thayer
et al., 1987). Pogson and Morris (1997) suggest that a possible reason for
induction of ethylene production may be maintenance of the reproductive
organs through mobilising metabolites from surrounding tissue, such as
sepals. Rather than maintaining floral organs post-harvest, it seems more
likely that broccoli is forced into premature senescence by the stresses at
harvest such as loss of water relations, lack of nutrients and cellular damage
(Pogson et al., 1995a). Once senescence is initiated from the stress-related
ethylene production, it continues in a genetically programmed way as during
normal senescence to support the fertilised embyro (seed) with nutrients. This
effect is exacerbated due to the high stress conditions of harvest and rapid
metabolism of nutrients from an ever decreasing pool.
25
1.10 Tomato fruit ripening and broccoli post-harvest senescence
It is difficult to understand the relationship between tomato fruit ripening
and broccoli post-harvest senescence without comparing and contrasting the
underlying mechanisms. It is not simply a case of comparing a tissue
undergoing ripening with one at the latter developmental stages undergoing
senescence. Broccoli is an immature plant at harvest that is forced into
premature senescence.
The onset of tomato fruit ripening is characterised by a climacteric
rapid increase in ethylene production. This increases the respiration rate
leading to the metabolism of macromolecules and a characteristic change
from green to orange/red colouration. Application of exogenous ethylene to
pre-climacteric fruit such as tomato hastens the ripening process. Makhlouf et
al. (1989) indicated that broccoli may be classified as climacteric, since florets
became yellow as respiration and ethylene production increased. It is clear
that ACC oxidase is up-regulated in both broccoli and tomato, correlating with
the level of ethylene. It appears that ACC synthase is also up-regulated in
tomato fruit ripening, and possibly in broccoli. Up-regulation of both these
genes results in a burst of ethylene leading to higher respiration rates and
metabolism of macromolecules (King and Morris, 1994a,b). Chlorophyll
molecules are broken down, starch hydrolysed into sugars, proteins and lipid
molecules are subsequently degraded. However, fruit ripening is
developmentally regulated so that the chloroplasts are converted to the
carotenoid rich chromoplasts to protect from oxidation products and the pool
of nutrients is not respired so rapidly. The harvested broccoli has no more
nutrient supply and quickly depletes available pools. It is characterized by a
26
change from chlorophyll-rich green colour to the yellow colour where carbon
based molecules have been oxidised leaving reduced nitrogen molecules,
and the accumulation of ammonia.
The capability of tissues to synthesise large quantities of ethylene in
response to ethylene is referred to as autocatalytic ethylene production (Yang
and Oetiker, 1998). It is characteristic of ripening fruits and senescing tissues
where ethylene is produced. It has been reported that the positive-feedback
response is due to the upregulation of ACC oxidases, and in tomato also ACC
synthases (Kasai, 1996; Cordes, 1989). Interestingly, it was reported that the
promoters of such genes contained highly conserved motifs (AGCCGCC)
known as GCC boxes. Ethylene-responsive element-binding proteins that
belong to a specific class of basic class leucine zippers are able to bind to
such regions and induce transcription of down-stream sequences (Hao et al.,
1998; Buttner and Singh, 1997). It is possible that this mechanism may
contribute to the climacteric increase of ethylene and post-harvest production
in broccoli.
Finally, broccoli senescence occurs in vegetative tissue so useful
comparisons with leaf senescence may be drawn. John et al. (1995)
introduced an ACC oxidase antisense construct into tomato, which inhibited
ACC oxidase expression and ethylene synthesis. This clearly delayed leaf
senescence by between 10 –14 days, but did not induce a sustained stay-
green phenotype as with some of the tomato fruit. John et al. (1995) suggest
that once senescence begins it proceeds normally and that ethylene may be a
significant factor in the timing of senescence.
27
1.11 Project Aims
The literature reports that ethylene has a major role in the rapid
senescence of broccoli, and that blocking the key enzymes of synthesis may
reduce the rate of yellowing. The cDNAs for the biosynthetic enzymes ACC
oxidase 1, ACC oxidase 2, and ACC synthase, have been isolated from
Brasssica oleracea var. italica (Shogun) (Pogson et al., 1995a,b).
The main aim of this project is to down-regulate the ACC oxidase and
ACC synthase enzymes in Brasssica oleracea var. italica with native cDNAs
inserted in the antisense orientation, and to up-regulate these enzymes with
sense constructs. This aim may be divided into three main objectives.
1). Produce ACC oxidase 1 and 2, and ACC synthase sense and
antisense constructs.
2). Transform suitable broccoli cultivar, and regenerate tissue into
whole plants with the possibility of producing marker-free plants.
3). Analyse transformed plants for ethylene production and factors
coupled with senescence such as chlorophyll levels.
If successful, it will be possible to determine the role of ethylene in
senescence of broccoli, and the importance of the ACC enzymes in this
process. The project may also produce an increased shelf-life phenotype of
broccoli through delayed sepal yellowing.
28
2.0 Construction of Brassica
oleracea ACC oxidase and ACC
synthase sense and antisense
gene cassettes
29
2.1 Introduction
Broccoli (Brassica oleracea L. var. italica) is a floral vegetable that
when harvested senesces rapidly from a chlorophyll-rich green hue to one
with pronounced sepal yellowing (Tian et al., 1994). Post-harvest senescence
is characterised by a complex breakdown and oxidation of carbohydrates,
proteins and lipids and loss of product quality. In an attempt to ensure a
‘quality’ product for the consumer, major retailers specify a product life of only
two days from delivery to ‘purchase before date’, and a further two days ‘best
consumed by’. This short post-harvest life results in significant waste for the
retailer, a cost borne by the consumer.
The gaseous hormone ethylene appears to play a major role in post-
harvest sepal yellowing as chlorophyll loss is associated with an increase in
floret ethylene synthesis (Tian et al., 1994). Ethylene biosynthetic inhibitors
such as aminoethoxyvinylglycine and inhibitors of ethylene action (silver ions)
have been shown to delay chlorophyll loss (Wang, 1977; Aharoni et al., 1985;
Clarke et al., 1994), whereas exogenous ethylene treatment enhances
chlorophyll loss (Tian et al., 1994).
The gene for the ethylene-forming enzyme in plants, 1-
aminocyclopropane-1-carboxylic acid (ACC) oxidase was first isolated from
tomato (Lycopersicon esculentum var. Ailsa Craig) by Hamiliton et al. (1990).
It was confirmed by down-regulating the gene with an inverted DNA sequence
thought to contain the gene. This ‘antisense’ approach not only determined
the presence of the gene in the DNA sequence but also provided a technique
for effective gene silencing. The DNA fragment from this work has enabled
the cloning of ACC oxidases in a number of crop species including Brassica
30
oleracea (Pogson et al., 1995a). The other major enzyme in the ethylene
biosynthetic pathway is 1-aminocyclopropane-1-carboxylic acid (ACC)
synthase that synthesises ACC from S-adenosylmethionine (Yang and
Hoffman, 1984) that is then oxidised by ACC oxidase to produce ethylene.
The cDNA from this gene was first cloned by Sato and Theologis (1989) from
zucchini fruits and has since been cloned from several other species.
Pogson and co-workers (1995a,b) used known cDNA probes from
mustard and apple to isolate two ACC oxidases and an ACC synthase cDNA
from B. oleracea L. cv Shogun. ACC oxidase 1 (aco1) and ACC oxidase 2
(aco2) shared nucleotide identity of 83% in putative translated regions and
48% in the 5’ untranslated regions. Post-harvest expression studies were
undertaken to establish the role of these genes in ethylene production and
their role in senescence. aco1 transcripts were found at low levels in the
whole florets and increased markedly in abundance after harvest. aco1
transcripts also increased in abundance in sepals after harvest. Transcripts
corresponding to aco2 were found exclusively within the reproductive
structures. They were absent at harvest but started to increase in abundance
within 2hr of harvest and accumulated to high levels. Pogson et al. (1995a)
concluded that aco1 was responsible for the basal level of ethylene required
for plant physiological and developmental processes, whereas aco2 was
specifically induced by stress conditions as experienced during harvest. ACC
synthase (acs), often thought to be the rate-limiting enzyme in the
biosynthesis of ethylene, showed no change in post-harvest transcript
abundance in the florets, and was detected at similar levels in green and
senescing leaves.
31
The main objective of this project is to test the hypothesis that post-
harvest production of ethylene is one of the major factors involved in broccoli
senescence. The ACC cDNAs isolated by Pogson et al. (1995a,b) were
obtained to down-regulate the native broccoli ACC genes, by in vivo gene
silencing as demonstrated by Hamilton et al. (1990). This approach may help
to determine specific roles for aco1, aco2 and acs in broccoli post-harvest
senescence and extend product shelf-life. A control plasmid with -
glucuronidase (gus) is necessary to follow changes of ethylene production in
transgenic plants without the ACC constructs.
Hamilton et al. (1990) showed that for effective gene silencing, high
levels of mRNA transcripts are important. They demonstrated the
effectiveness of the cauliflower mosaic virus (CaMV) 35S promoter in
producing high levels of antisense ACC oxidase mRNA transcripts in tomato,
which resulted in a reduction of ethylene production. The CaMV 35S promoter
is active in the genome of a variety of monocotyledonous and dicotyledonous
plants (Blaich, 1992). Benfey and Chua (1990) determined that the CaMV 35S
promoter comprises a collection of sequences that individually direct a defined
pattern of tissue specific and developmentally regulated gene expression
which cumulatively result in constitutive gene expression. It is a suitable
promoter for studying the down-regulation of the ACC genes in different
tisssues and stages of broccoli development, especially post-harvest. The
nopaline synthase (nos) terminator (Bevan, 1983) has a polyadenylation
signal. Polyadenylation of mRNAs is thought to confer greater stability to the
mRNA and increase efficacy of nucleic acid interaction required for gene
silencing.
32
It is advantageous to produce marker-free plants that are
phenotypically normal and do not contain superfluous bacterial DNA. A binary
vector such as pSCV1.0 (Biogemma) contains a minimal amount of T-DNA
between the borders without a selectable marker. There are eleven unique
restriction sites within the T-DNA borders that can be used to clone genes of
interest. Outside the borders the bacterial gentamicin/kanamycin and the pUC
ampicillin ( -lactamase, bla) resistance genes are present. pSCV1.0 also
contains co-existent origins of replication (Col E1 and RK2) that produce high
copy numbers in both E. coli and Agrobacterium.
The aim for this part of the project is to clone the ACC cDNAs into the
minimal T-DNA binary vector pSCV1.0 in both sense and antisense
orientations between the CaMV 35S promoter and the nos terminator.
33
Sst
RB CaMV 35S m-gfp5-ER
2.2 Materials and Methods 2.2.1 Cloning Strategy
(a). The gus gene was removed from pBI121 with XbaI and SstI restriction
enzymes.
RB CaMV 35S gus nos LB
HindIII XbaI SstI EcoRI
(b). The gus gene was replaced with an oligo-nucleotide (linker) which had been designed with suitable restriction endonuclease restriction sites.
XbaI BamHI SalI SmaI SstI
CaMV 35S
TCTAGAGGATCCATGTCGACCCGGGAGCTC AGATCTCCTAGGTACAGCTGGGCCCTCGAG
nos
HindIII EcoRI
(c). The HindIII/EcoRI cassette was transferred into the multiple cloning site of the pSCV1.0 and renamed pJ1.0
The Multiple Cloning Site of pSCV1.0
BglII BclIHindIII PstI SalI XbaI BamHI SmaI XmaI EcoRIEcoRV
CaMV 35S linker nos
HindIII EcoRI
34
T-DNA region of pJ1.0 (All sites unique except those in bold)
RB BglII, BclI, HindIII, SphI, PstI CaMV X 35S
baI, BamHI, SalI, SmaI, SstI nos EcoRI, EcoRV LB
(d). aco1 and aco2 were PCR-amplified from pBluescript and the products
were phosphorylated with the pMOS blue kit. pJ1.0 was digested with SmaI
and the cut-ends were dephosphorylated. The cDNA clones were ligated into
pJ1.0.
SmaI
P RB CaMV 35S
aco1 and aco2
SmaI
P nos LB
(e). acs was PCR-amplified from pBluescript with primers adding BamHI sites
to the ends. Both acs and pJ1.0 were digested with BamHI and then ligated
together.
BamHI BamHI
RB CaMV 35S acs
nos LB
(f). For the control plasmid, the HindIII/EcoRI fragment from pBI121 was
ligated into the HindIII/EcoRI restriction sites of pSCV1.0.
2.2.2 Plasmids 2.2.2.1 The ACC oxidase and ACC synthase cDNAs
The B. oleracea ACC oxidase and ACC synthase cDNAs (Pogson et
al., 1995a,b) were obtained from Kevin Davies (Crop and Food Research,
35
Levin, New Zealand). The genes had been ligated into the EcoRI site of
pBluescript SK, after attachment of NotI, EcoRI adapters, which had been
checked by sequencing with the T3 and T7 promoter primers. The sequence
analysis showed the adaptor sequence, and presence of the gene within the
multiple cloning site of pBluescript. The aco1 gene had been ligated in the
sense orientation, whereas the aco2 and ACC synthase sequences were in
antisense orientation relative to the T7 promoter.
2.2.2.2 Binary Vectors
The binary vector pSCV1.0 (Figure 2.2; Biogemma, Cambridge) was
used as the backbone for the constructs. The binary vector pBI121
constructed by Jefferson et al. (1987) (Figure 2.1) and obtained from Dr J.
Gittins, (HRI-East Malling, Kent) was used as a source of promoters and
terminators as well as the -glucuronidase (gus) gene for a control. pBI121 is
derived from pBin19 (Bevan, 1984). It contains the CaMV 35S promoter
fragment (Guilley et al. 1982), the -glucuronidase gene (Jefferson et al.,
1987) and the nos terminator (Bevan et al., 1983) within the T-DNA region. pBI121 also has the neomycin phosphotransferase (nptII) gene which
detoxifies a number of aminoglycoside antibiotics, including neomycin,
kanamycin and geneticin.
2.2.2.3 Plasmid extraction
All plasmids were extracted with Qiagen miniprep (tip20) kits (Qiagen
Ltd., Boundary Road, Crawley, West Sussex), which produced sufficient DNA
for the next step of the cloning process.
36
nos-p nptII (kanR) nos-t CaMV 35S gus nos-t
HindIII
SphI
EcoRV
XbaI
BamHI
SstI
EcoRI
RB LB
T-DNA
pUC Ori
pBI121
nptIII RK2
Figure 2.1 Circular map of pBI121 (adapted from Frisch et al., 1995 and Jefferson et al., 1987) showing the T- DNA region with unique restrcition endonuclease sites.
37
Unique cloning sites between T-DNA borders:
EcoRV, EcoRI, XmaI, SmaI, BamHI, XbaI, SalI, PstI, HindIII, BclI, BglII
ClaI
Left
Right
SstI ClaI
border unique cloning sites border OD
T-DNA
ScaI
(8.05)
AmpR
AccI
(2.28)
Col E1
SphI
(5.85)
pSCV1.0
(8.5kb)
Gm/KmR RK2
ScaI
(5.2)
SstI
(4.98)
KpnI
(4.9)
SphI
(3.23)
Figure 2.2 Circular map of pSCV1.0 It shows the ampicillin, kanamycin/gentamycin bacterial resistance genes and origins of replication. Within the T-DNA borders, the eleven unique restriction endonuclease sites of the multiple cloning site are shown which can be used for cloning genes of interest.
38
2.2.3 Bacterial strains and media
All cloning steps were conducted in E. coli DH5 , genotype: F-
80dlacZ M15 (lacZYA-argF)U169 deoR recA1 endA1 hsdR17(rk
-, mk+)
phoA supE44 -thi-1 gyrA96 relA1 (GIBCO-Life Technologies, Uxbridge, UK.)
and cultured in LB media (see appendix). Transformed cells with either
pBI121 or pSCV1.0 derivatives were selected for with 25mg/l kanamycin and
those with pBluescript, 50mg/l ampicillin. E. coli cells were incubated at 37oC.
Agrobacterium rhizogenes strain LBA 9402 pRi1855::pMBRE36GFP
was cultured either on YMB plates, or MGL (appendix) liquid broth at 25oC.
The cells were selected for with tetracyclin, rifampicin and kanamycin for the
cointegrate Ri plasmid, chromosomal DNA and binary plasmid, respectively.
2.2.4 Enzymes and DNA
All enzymes including restriction enzymes, DNA ligases, and Taq
polymerases were purchased from GIBCO-Life Technologies, Uxbridge, U.K
as were DNA ladders for gel electrophoresis. All oligonucleotides were
synthesised by VH Bio Ltd., Gosforth, Newcastle upon tyne, UK (Table 2.1)
and used in standard PCR conditions (Table 2.2).
39
Table 2.1 Oligonucleotide sequences used in these experiments
Primer
No. Name Sequence (5’- 3’) Tm
(a)
(oC)
1 Oligo top CTAGAGGATCCATGTCGACCCGGGAGCT 72
2 Oligo bot CCCGGGTCGACATGGATCCT 67
3 35S (F) ACGTTCCAACCACGTCTTCA 60
4 nos (R) CCGATCTAGTAACATAGATGACAC 61
5 aco1 (F) TCAAACAACTAGCTACTTCAAC 57
6 aco1 (R) TTTCTTAGAACAGTTATCATATTATTT 54
7 aco2 (F) TCATATTTCCAATACAATTATTGC 54
8 aco2 (R) CTAACATCATCAACAACATTAAG 56
9 acs (F) GCGGGATCCAATAACGATTAGAGAAATTATGGT 69
10 acs (R) GCGGGATCCATCTCAACAGAACAAAAACAAAC 67
11 aco1(I) ACGTTCCAACCACGTCTTCA 60
12 aco1 (II) AGCTTCATCGCTTGCCTGTA 60
13 aco1 (III) CTTCTCGACGTTGTCCATCA 60
14 aco1 (IV) CCGATCTAGTAACATAGATGACAC 61
15 aco2 (I) GACATCTCCACTGACGTAAG 60
16 aco2 (II) CCACAGTCGAGACGTTCTAA 60
17 aco2 (III) TTCATGGCCGCTCGGTATTC 62
18 aco2 (IV) CCGATCTAGTAACATAGATGACAC 61
19 acs (I) GACATCTCCACTGACGTAAG 60
20 acs (II) AGAAGTGTTGGCAGAGTAGC 60
21 acs (III) CTTCGCTACAGCTTGTCTGA 60
22 acs (IV) CCGATCTAGTAACATAGATGACAC 61
(a) Tm of the primers was calculated with the expression:
Tm = 81.5 + -16.6 + (41 x (#G + #C/length)) -(500/length)
40
Table 2.2 PCR reaction components. To make a total 25 l for each reaction, 5 l of template DNA (100pmol/ l) and 5 l of primers (set at 4pmol/ l) were added. Primers and template DNA were suspended in autoclaved reverse osmosis water.
x1 ( l)
10x Buffer 2.5
50mM MgCl2 0.75
dNTPs 25mM 0.2
H2O 11.42
BRL Taq 0.13
Total 15
The PCR program in Table 2.3 was used and annealing temperatures
adjusted to 3oC below Tm of primer pairs (Table 2.1). This was performed on a
Hybaid Omnigene PCR machine.
Table 2.3 PCR program. Stage 1 and 3 were cycled once. Stage 2 was cycled thirty five times.
Stage 1. 93oC/ 2 min 58o C/ 1 min
Stage 2. 72o C/ 1 min 93oC/ 30 s 58oC/ 1 min
Stage 3. 72oC/ 10 min
41
T
2.2.5 Cloning Strategy 2.2.5.1 Removing gus from pBI121
pBI121 was extracted with a Qiagen miniprep and 1 g was digested
with XbaI and SstI (5 units each) in the buffers specified in 20 l autoclaved
reverse osmosis water for 2h at 37oC. The digest products were loaded, with
bromophenol blue dye, onto a 0.7% (w/v) agarose gel containing 0.33 g/ l
ethidium bromide. The Low DNA mass ladder (GIBCO-Life Technologies,
Uxbridge, UK) was loaded alongside these samples. The gel was run for 50
minutes at 80 Volts, 0.1 amps in a BioRad gel tank. The digested pBI121
separated into 2 bands: one at 2Kbp, the gus gene; and the other 11Kbp, the
remaining backbone. The 11Kbp fragment was cut out of the gel with a clean
scalpel blade and placed into a 1.5ml microfuge tube. The DNA was purified
with the Qiaex II kit (Qiagen Ltd., Boundary Road, Crawley, West Sussex),
yielding 150ng of cut plasmid.
2.2.5.2 Replacing gus with synthetic linker
A synthetic DNA linker was created by annealing oligonucleotides 1
and 2. They were initially heated to 75oC in a water bath and left to cool for
2h. The linker contained 5 restriction endonuclease sites as shown below:
XbaI BamHI SalI SmaI SstI
CTAGAGGATCCATGTCGACCCGGGAGC TCCTCGGTACAGCTGGGCCC
The linker was ligated into the XbaI/SstI sites of the cut pBI121
plasmid. The digested pBI121 plasmid was 11Kbp and the synthetic linker
42
24bp in length and therefore, a 1:5 ratio by number would be 11000: 120 by
weight. For the ligation, 45ng of plasmid and 0.45ng linker was used. The
annealed synthetic linker was diluted to a concentration 0.1ng/ l. The
digested plasmid was at a concentration 7.5ng/ l. The ligation reaction
mixture consisted of: H2O 5.4 l; digested plasmid 6.0 l; linker 4.5 l; T4
ligase buffer 4.0 l and T4 DNA ligase 0.1 l, making a total 20 l. The ligation
mixture was incubated at 22oC for two hours.
The transformation of ligated plasmid was carried out according to the protocol supplied with the DH5 library efficient competent cells (GIBCO-Life
Technologies). The cells transformed with the binary plasmid were selected
with 25mg/l kanamycin on LB agar plates and grown overnight at 37oC. Single
colonies were checked for presence of the synthetic linker insert with a colony
PCR method taken from the pMOSBlue T-vector kit (Amersham International
plc, Little Chalfont, Buckinghamshire, UK). This involved picking a white
colony, approximately 1mm in diameter from the plate using a sterile cocktail
stick. The tip was touched onto a master plate and then transferred to a 1.5ml
tube containing 50 l sterile water. This was repeated for ten colonies. The
tubes were placed in boiling water for 2 min to lyse the cells and denature
genomic DNA and nucleases. The tubes were put on ice for a further 2 min for
the plasmid DNA to renature and then centrifuged at 12000g for 1 minute to
remove cell debris. In the PCR, the linker oligonucleotides (1 and 2) were
used as primers, paired with primers designed from the CaMV 35S (3) and
nos (4) sequences (Table 2.1). Colonies with positive PCR bands were grown
overnight at 37oC in liquid LB and the plasmids were extracted with the
Qiagen miniprep kit. Plasmid integrity was checked with restriction digests.
43
2.2.5.3 Transferring CaMV 35S-linker-nos cassette from pBI121 to pSCV1.0
E. coli DH5 cells containing the pSCV1.0 plasmid were cultured
overnight and plasmids extracted with the Qiagen miniprep tip20 kit. The
pSCV1.0 and the pBI121 plasmid with the linker instead of gus were digested
with EcoRI and HindIII with the specified buffers in 20 l sterile R.O. water at
37o C for 2h. The digested products were run on a 0.7% (w/v) agarose gel at 80 volts for 50 minutes. The 1.1Kbp fragment from the adaptered pBI121
plasmid and the 8.5 Kbp pSCV1.0 backbone were cut out of the gel with a
clean scalpel and placed into 1.5 ml microfuge tubes. The DNA was re-
extracted with the Qiaex II kit, supplied by Qiagen and quantified on an
agarose gel in preparation for ligation. The EcoRI/HindIII fragment from
pBI121 yielded a concentration of 1.5ng/ l, from an 18 l sample. The
digested pSCV1.0 backbone yielded 75ng/ l, from a sample of 20 l. The size
of pSCV1.0 was 8.5Kbp, whereas the EcoRI/HindIII fragment from the
pBI121 plasmid was 1.1Kbp. Therefore, a 1:1 ratio by number would be
8.5:1.1 by weight and a 1:5 ratio 8.5:5.5. As the total quantity of fragment DNA was 27ng in 18 l, the amount of pSCV1.0 plasmid DNA required for a
1:5 ratio would be 42ng. As the concentration of pSCV1.0 was 75ng/ l, only
0.55 l of DNA was required for the ligation. The ligation reaction consisted of:
18 l EcoRI/HindIII fragment; 0.6 l EcoRI/HindIII pSCV1.0 digest; 4 l T4
DNA ligase buffer; 0.1 l T4 DNA ligase; total 22.7 l. The ligation was carried
out at 22oC overnight. DH5 cells were transformed with the ligation mixture
according to the GIBCO-Life Technologies protocol, and transformants were
selected for on LB agar plates with 25mg/l kanamycin. Single colonies were
checked with the PCR method described above with primer pairs of
44
oligonulceotides 1 and 4 and 3 and 2 (Table 2.1). Colonies giving bands were
cultured in an overnight LB broth at 37oC. The plasmids were extracted with
the Qiagen miniprep kit and checked by digestion with restriction
endonucleases (Figure 2.4b) and called pJ1.0.
2.2.5.4 Inserting the Brassica oleracea ACC oxidase 1 and 2 cDNAs into pJ1.0 by blunt-ended cloning
The ACC oxidase genes were PCR amplified from pBluescript with
primers 5-8 (Table 2.1). As the synthetic primers flank either end of the ACC
cDNAs, bands of 1231bp for aco1 and 1223 bp for aco2 were expected
(Figure 2.5). The PCR products were separated from the pBluescript template
DNA on a 0.7% (w/v) agarose gel run for 30min at 100 volts. The PCR
products were cut out of the gel with a clean scalpel and placed into a 1.5ml
microfuge tube. The DNA was re-extracted from the agarose gel with the
Qiaex II kit (Qiagen). This yielded 250ng/ l for each ACC oxidase cDNA
clone.
Five micrograms of the pSCV1.0 35S-linker-nos plasmid was digested
with SmaI as specified by GIBCO-Life Technologies for 2h at 30oC. As SmaI
produces a blunt-ended fragment it was essential to dephosphorylate the cut
ends to prevent the plasmid from closing up without insert during the ligation.
After the digest, the incubation temperature was raised to 37oC and 1.59 l of
calf intestinal alkaline phosphatase (CIP) was added. This figure was arrived
at by a simple set of calculations. The molecular weight of the plasmid was
roughly 9500bp X 660 (molecular weight of a pair of nucleotides) = 6270000 =
1gmole.
45
1g = 1 = 1.6 X 10-7 moles 6270000
1 g = 1 = 1.6 X 10-13
6270000 X 106
1 g = 1 X 1012 = 0.16 pmoles 6270000 X 106
1 g = 0.32 pmole ends (2 ends per molecule)
As there were 5 g of plasmid and 1 pmol ends requires 1 unit of
enzyme, 1.6 units of enzyme (i.e. 1.6 l as 1u/ l) were added and incubated
for 1 h. The CIP was promptly inactivated at 75oC for ten minutes and DNA
repurified with a phenol:chloroform extraction as below. The digested and
dephosphorylated plasmid was made up to 100 l by adding 80 l R.O. water.
An equal volume of phenol (100 l) was added and the two were mixed by
inverting the tube 5 times. The tube was spun in a microfuge at 11600g for ten
seconds. The aqueous phase was removed and placed into another tube
while the phenol was carefully discarded. Both phenol and chloroform were
added to the aqueous phase in a 50:50 ratio and mixed and separated as
above. Chloroform was added (100 l) and again repeated as above. The
aqueous phase was removed and to this 230 l ethanol was dispensed and
centrifuged at 11600g for 20min. The supernatant was removed and the DNA
pellet air dried for 5 min. The pellet was resuspended in 20 l sterile R.O.
water. The DNA was quantified and the yield found to be 100ng/ l.
46
2.2.5.5 pMOSBlue kit
To ligate the ACC oxidase 1 and 2 PCR products into pJ1.0, the
fragments were blunt-ended and phosphorylated. This was done using
components from the pMOSBlue blunt-ended cloning kit (Amersham Life
Science). The pSCV1.0 35S-linker-nos plasmid was 9.5Kbp and the cDNA
insert 1.3Kbp. Therefore, a 1:2.5 ratio (9.5:3.25) would be 50ng plasmid and
17ng insert. This quantity of insert was added to the pMOSBlue reaction
mixture as specified in the protocol and incubated for 40 min at 22oC. The
reaction was promptly heat inactivated at 75oC for ten minutes and placed on
ice for 2 min. For the ligation, 50ng of SmaI cut plasmid and 1 l T4 DNA
ligase (4 Weiss units) were added and incubated overnight at 15oC.
Transformation with DH5 cells and 1 l ligation reaction mixture was carried
out according to the GIBCO-Life Technologies’ protocol that comes with the cells and transformants selected using 25mg/l kanamycin in LB agar plates.
2.2.5.6 Inserting Brassica oleracea ACC synthase cDNAs into pJ1.0
The ACC synthase gene was PCR amplified from pBluescript with
primers 9 and 10 flanking both ends of the cDNA clone with BamHI sites at
the 5’ ends. The synthetic primers produced a band 1746bp including the
cDNA and BamHI sites (Figure 2.5). The PCR products were separated from
the pBluescript template DNA on a 0.7% (w/v) agarose gel run for 30min at
100 volts. The PCR products were cut out of the gel with a clean scalpel and
placed into a 1.5ml microfuge tube. The DNA was re-extracted from the
agarose gel with the Qiaex II kit (Qiagen) and yielded 300ng/ l for 20 l. The
product was digested with BamHI (GIBCO-Life Technologies) with the correct
47
buffer at 37oC for 2h. The restriction enzyme was inactivated at 80oC for 20min and then cooled on ice for 2min. Propan-2-ol (400 l) was added to the
tube and spun in a microfuge at 11600g for 5 min to precipitate the fragment
without the small pieces from the cut ends. The supernatant was removed
and 0.5ml 70% ethanol was added. This was spun in a microfuge as before.
The supernatant was removed and the DNA pellet air dried for 5 min. The
pellet was resupended in 20 l sterile R.O. water and adjusted to 200ng/ l.
Five micrograms of the pSCV1.0 35S-linker-nos plasmid was digested
with BamHI (GIBCO-Life Technologies) in the correct buffer at 37oC for 2h. It
was dephosphorylated and repurified as described above with the SmaI cut
plasmid. This yielded 100ng/ l.
The pSCV1.0 35S-linker-nos plasmid was 9.5Kbp and the cDNA insert 1.738Kbp. Therefore, a 1:5 ratio would be 9.5:8.7, or 50ng plasmid, 45ng
insert. The ligation reaction mixture comprised of 0.5 l plasmid, 0.23 l insert,
4.0 l T4 DNA ligase buffer (GIBCO-Life Technologies), 0.1 l T4 DNA ligase (GIBCO-Life Technologies) and 15.2 l sterile R.O water. The ligation was
carried out overnight at 22oC. The transformation was performed according to
the protocol provided by GIBCO-Life Technologies, with the competent DH5
cells. Transformed cells were selected on 25mg/l kanamycin LB agar plates.
2.2.5.7 Checking transformants for inserts
Twenty colonies from each of the ACC oxidase 1 and 2 and ACC
synthase plates were analysed with the colony PCR method described earlier.
The same primer pairs as those used to amplify the cDNA from pBluescript
were used to check presence of an insert. The DNA from the colonies with
48
positive bands on the gel was tested with four primer pairs to determine
orientation. Figure 2.3 shows a schematic diagram of the orientation of the
primer pairs depending on insertion of the cDNA. For example, if there has
been a aco1 cDNA inserted in antisense orientation, primers 3 and 4 will
produce a 542bp DNA fragment, whereas primers 3 and 5 will not produce a
fragment as in this orientation they are both forward primers (band sizes are
shown in Table 2.4).
Table 2.4 Predicted PCR product sizes for the insertion of ACC cDNAs in sense or antisense orientations into pJ1.0.
Construct Orientation Primer Pairs Predicted fragment size
aco1 antisense 3+4 542 aco1 antisense 5+nos 462 aco1 sense 3+5 360 aco1 sense 4+nos 710 aco2 antisense 3+8 369 aco2 antisense 9+nos 676 aco2 sense 3+9 578 aco2 sense 9+nos 465 acs antisense 3+12 471 acs antisense 13+nos 628 acs sense 3+13 530 acs sense 12+nos 572
49
a RB 3
5 LB
R CaMV 35S aco1 antisense nos
4 nos
b RB 3 4 LB
R CaMV 35S aco1 sense nos
5 nos
c RB 3 9 LB
R CaMV 35S aco2 antisense nos
8 nos
d RB 3 8 LB
R CaMV 35S aco2 sense nos
9 nos
e RB 3
13 LB
R CaMV 35S acs antisense nos
12 nos
f RB 3
12 LB
R CaMV 35S acs sense nos
13 nos
Figure 2.3 Schematic diagram of primer pairings for cDNA inserts in antisense and sense orientations for aco1, aco2 and acs.
50
The plasmids with positive bands were also extracted with the Qiagen
miniprep tip20 kit and digested with BglII (GIBCO-Life Technologies) to check
orientation of insert and plasmid size (Figure 2.7a of results).
2.2.5.8 Control Plasmid
A control plasmid was constructed with the pSCV1.0 backbone and
pBI121 CaMV 35S-gus-nos gene cassette. DH5 cells containing the
plasmids were selected for with 25mg/l kanamycin and cultured overnight in
liquid LB at 37oC in a shaking incubator. The plasmids were extracted with a
Qiagen miniprep tip20 kit and 5 g restriction digested with EcoRI/HindIII
(GIBCO-Life Technologies) with the specified buffers at 37oC for 2h. The
digested products were run on a 0.7% (w/v) agarose gel at 80 volts for 50
minutes. The pSCV1.0 backbone (8.5Kbp) and the pBI121 CaMV 35S-gus-
nos cassette (3Kbp) were cut out of the gel with a clean scalpel and placed
into a 1.5ml microfuge tube. The DNA was re-extracted with the Qiaex II gel
purification kit (Qiagen) and quantified. The CaMV 35S-gus-nos cassette
yielded 10ng/ l (200ng/20 l) and pSCV1.0 backbone 150ng/ l (3 g/20 l).
The size of the pSCV1.0 backbone was 8.5Kbp, and the gene cassette 3Kbp.
For a 1:5 plasmid:insert ratio, 50ng backbone plasmid and 90ng gene
cassette were used for the ligation. The ligation mixture comprised of: 9 l
insert, 0.33 l backbone plasmid, 4 l T4 DNA ligase buffer, 0.1 l T4 DNA
ligase and 6.57 l sterile R.O. water. The ligation mixture was incubated for 2h
at 22oC. The transformation was performed according to the protocol provided
by GIBCO-Life Technologies, with the competent DH5 cells. Transformed
cells were selected on 25mg/l kanamycin LB agar plates. Transformants were
51
checked with the colony PCR method described earlier with primers 11 and 13 (Table 2.1) and the plasmids were digested with EcoRI /HindIII.
2.2.5.9 Producing and storing E. coli and A. rhizogenes stocks with the 7 constructs
The E. coli DH5 cells with the constructs were cultured on liquid LB
with 25mg/ml kanamycin and grown overnight at 37oC in a shaking incubator.
They were grown to a density of approximately 1 x 109 cells/ml. Sterile
glycerol was mixed 50:50 with liquid LB. Into a sterile screw cap 1.5ml tube,
0.5ml DH5 cells and 0.5ml glycerol/LB was added and gently mixed by
inverting. These were frozen at –80oC and stored for future use.
Cells from A. rhizogenes strain LBA 9402 virulence plasmid pRi1855,
cointegrate with pMBRE36 GFP were rendered competent using the
technique from Mattanovich et al. (1989). Competent cells were transformed
by electroporation (Mattanovich et al., 1989) using a Biorad Gene Pulser.
Transformants were selected for on YMB plates with 50mg/l kanamycin,
100mg/l rifampicin and 10mg/l tetracyclin and grown for 3 days at 25oC. They were stored in the same way as the DH5 cells but with MGL instead of LB
liquid broth.
52
Figure
2.3. Results
The backbone plasmid, pJ1.0 was constructed to be used as a cloning
vector for the ACC cDNAs. It was based on pSCV1.0 with the addition of the
CaMV 35S promoter and nos terminator, flanking a multiple cloning site linker.
It is shown schematically in Figure 2.4 (a), and confirmed in Figure 2.4(b).
(a) HindIII
RB
BamHI SmaI EcoRI
LB
CaMV 35S Promoter Linker nos
888bp 277bp
876bp 289bp
1165bp
(b)
bp
2000 1200 800
400 200
1 2 3 4 5 6 7
Figure 2.4 (a) T-DNA region of pJ1.0 with restriction sites. (b) Restriction endonuclease digest confirming the integrity of the backbone plasmid, pJ1.0 The plasmid was digested with the restriction enzymes shown in Figure 2.4a. Lane 1, DNA Mass Ladder; Lane 2, uncut plasmid, Lane 3, EcoRI/HindIII; Lane 4, EcoRI/HindIII/SmaI; Lane 5, EcoRI/HindIII/ BamHI; Lane 6, SmaI; Lane 7, BamHI.
53
In Figure 2.4 (b), lane 3, the EcoRI/HindIII digest produced an 1165bp
fragment. These sites are within the T-DNA borders and the fragment
includes the CaMV 35S promoter, the linker, and the nos terminator. In lanes
4 and 5 the fragment including the promoter, linker and terminator is cut at the
linker multiple cloning site by SmaI and BamHI, respectively. The band seen
on the gel at 870bp is the 35S promoter. There should also be a nos
terminator band (270bp), but there is not enough DNA for it to be visible.
Although faint bands around 450bp are present these are unidentified, as the
nos bands would run further. However, the difference in sizes of the bands in
lanes 4 and 5 to the band in lane 3 is the size of the nos terminator.
The major bands in lanes 6 and 7 show linearised plasmids with just a
single cut. They show that SmaI and BamHI sites within the linker must be
unique within pJ1.0 and can be used to clone the ACC broccoli cDNAs.
Following cloning of ACC cDNA inserts into the pJ1.0 linker multiple
cloning site, colony PCR was used to confirm the success of the procedure.
The cDNA clones were originally amplified from pBluescript (SK) with specific
primers. The same primers were used to check the transformed E. coli
colonies produced from a ligation of vector and insert. A colony producing a
positive band was then analysed with pairs of primers to determine the
orientation of the insert.
In Figure 2.5, the aco1 and aco2 inserts were 1232/1237bp,
respectively, and the ACC synthase 1700bp in size (lanes 2-4). The presence
of these bands demonstrates the insertion of the aco1, aco2 and acs cDNA
clones. As the cDNAs were cloned non-directionally, they could insert in either
sense or antisense orientation. Therefore, once it was clear an insert was
54
present, a second PCR was conducted to determine orientation. In order to
avoid inaccuracies, the technique, detailed in the methods was based on a
band/no band result. This worked very effectively for aco1. Lanes 5-8 show
aco1 antisense, and lanes 9-12 aco1 sense.
bp
1353 1078 872 603 310
1 2 3 4 5 6 7 8 9 10 11 12
Figure 2.5 PCR products to determine presence and orientation of insert (see Figure 2.3/Table 2.4). In lane 1, x174/HaeIII digest (marker);
lane 2, aco1 cDNA insert in pJ1.0; lane 3, aco2 cDNA insert in pJ1.0; lane 4, acs cDNA insert in pJ1.0; lane 5, aco1 A (primers 3+4); lane 6, aco1 A (3+5); lane 7, aco1 A (5+nos); lane 8, aco1 A (4+nos); lane 9, aco1 S (3+4); lane 10, aco1 S (3+5); lane 11, aco1 S (4+nos); lane 12, aco1 S (5+nos).
55
bp
1353 1078
872 603
310
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Figure 2.6 PCR products to determine orientation of aco2 and acs inserts. In lane 1, x174/HaeIII digest markers; lane 2, aco2 A (3+8); lane 3, aco2 A
(3+9); lane 4, aco2 A (9+nos); lane 5, aco2 A (8+nos); lane 6, aco2 S (3+8); lane 7, aco2 S (3+9); lane 8, aco2 S (9+nos); lane 9, aco2 S (8+nos); lane 10, acs A (3+12); lane 11, acs A (3+13); lane 12, acs A (13+nos); lane 13, acs A (12+nos); lane 14, acs S (3+12); lane 15, acs S (3+13); lane 16, acs S (13+nos); lane 17, acs S (12+nos).
In Figure 2.6, lanes 2 and 4 show the predicted bands for aco2
antisense (as shown in Table 2.4). However, bands were also present in
lanes 3 and 5. The predicted bands for aco2 sense for these lanes was 578bp
and 465bp and is therefore indistinguishable. It was the same for aco2 sense,
indicating that sequences elsewhere on the plasmid had been amplified non-
specifically. This made it impossible to determine the orientation of aco2 using
these primers.
ACC synthase sense and antisense orientations are shown in lanes
10-17. acs antisense exhibited the band/no band result that was expected
with lanes 10 and 12 showing bands of 471bp and 628bp, respectively.
Although the acs sense primers produced bands similar to acs antisense,
56
primers 3 and 13 and, 12 and nos (lanes 15 and 17) did not. The actual
results were not as straight forward as predicted but the insert could only be in
one of two orientations. Therefore, these results were used relative to each
other to determine that the inserts were in different orientations. Confirmation
of orientation was completed with a series of restriction digests.
(a)
bp 2000 1200 800
400 200
1 2 3 4 5 6 7 8
(b)
BglII RB
1781bp
1250bp
A S BglII BglII
LB
R CaMV 35S aco1 nos
(c)
BglII RB
1704bp
1281bp
A S
BglII BglII
LB
R CaMV 35S aco2 nos
57
(d)
BglII RB
1805bp
1633bp
S A
BglII BglII
LB
R CaMV 35S acs nos
(e) HindIII EcoRI
RB 870bp
1808bp 270bp LB
R CaMV 35S gus nos
Figure 2.7 Confirmation of cDNA inserts (a) Restriction digest to confirm orientation of insert and presence of promoter. In lane 1, DNA Mass Ladder; lane 2, aco1 antisense BglII; lane 3, aco1 sense BglII; lane 4, aco2 antisense BglII; lane 5, aco2 sense BglII; lane 6, acs antisense BglII; lane 7, acs sense BglII; lane 8 pJGUS EcoRI/HindIII. (b,c and d) aco1, aco2 and acs constructs depicting the size of band produced from BglII digests, depending on orientation of the insert. (e) pJGUS construct depicting the EcoRI/HindIII fragment produced in lane 8. RB=right border, LB=left border, A= antisense, S=sense.
In pJ1.0 there is a single BglII site within the right border. There are
also single BglII sites in the ACC cDNAs at asymmetrical positions along the
DNA. Therefore, if an insert is present, the two sites produce a band size
dependent on orientation. For example, in Figure 2.7 (a), lane 2 an aco1
antisense insert must be present as the 1250bp fragment includes 870bp of
the CaMV 35S promoter and 380bp of the aco1 cDNA. If the insert was in the
sense orientation as in lane 3, the band would be 1781bp including 870bp of
the CaMV 35S promoter and 910bp of the cDNA. This is diagrammatically
shown in Figure 2.7 (b). The same premise was used for aco2 and acs. If
aco2 was in antisense orientation it would produce a band 1281bp (lane 4)
and sense 1704bp (lane 5) as shown in Figure 2.7 (c). ACC synthase is
slightly different in that the antisense band is larger (1805bp) than the sense
58
band (1633bp) due to restriction site position. This can be seen in Figure 2.7
(a) lanes 6 and 7 and Figure 2.7 (d). In Figure 2.7 (a) lane 8, the pJGUS
control plasmid was digested with EcoRI/HindIII. The 2.9Kbp band shown
includes the CaMV 35S promoter (870bp), the -glucuronidase (1808bp) and
the nos terminator (270bp). This is diagrammatically shown in Figure 2.7 (e).
The BglII digests in lanes 2-7 also confirm the presence of the CaMV 35S
promoter which had been cloned in an earlier step.
Having determined the presence of the CaMV 35S promoter and
orientation of the ACC cDNA inserts in Figures 2.4 and 2.7, it seems
necessary to confirm the presence of the nos terminator. Due to its small size,
the nos band was expected to show little fluorescence and be difficult to see.
However, the cumulative addition of the nos terminator to a larger fragment
showed its presence. In Figure 2.8 (a), lanes 2-5 show BamHI/EcoRI digests
for aco1 and aco2. The digest produced 1570bp fragments, which are
diagrammatically shown in Figure 2.8 (b). The EcoRI site is to the right of the
nos terminator and the BamHI site to the left of the cDNA. The 1300bp cDNA
and the 270bp nos terminator produce a 1570bp as can be observed in Figure
2.8 (a).
59
1550bp
CaMV 35S cDNA nos
1738bp
CaMV 35S cDNA nos
(a)
bp
2000
1200
800
1 2 3 4 5 6
(b) aco1 and aco2
BamHI EcoRI
RB LB
R
(c) acs BamHI BamHI
RB LB
R
Figure 2.8 Confirmation of cDNA inserts and nos terminator (a) Restriction digest to confirm presence of insert and nos terminator. In lane 1, DNA Mass ladder; lane 2, aco1 antisense BamHI/EcoRI; lane 3, aco1 sense BamHI/EcoRI; lane 4, aco2 antisense BamHI/EcoRI; lane 5, aco2 sense BamHI/EcoRI; lane 6, Syn antisense BamHI; Lane 7, acs sense BamHI. (b) Schematic diagram of the fragment size that a BamHI/EcoRI
60
restriction digest of aco1 and aco2 constructs would produce. (c) Schematic diagram of the fragment size that a BamHI restriction digestion would produce from an acs construct with the insert.
Due to limitations with restriction sites, the BamHI digest of ACC
synthase in lanes 6 and 7 of Figure 2.8 (a,c) just show the presence of the
cDNAs (1700bp) and that both BamHI sites are intact from cloning.
Initially, the T-DNA region of pJ1.0 was sequenced to confirm the
presence of the CaMV 35S promoter, nos terminator and the left and right
borders. The primers used for this process were the oligonucleotides (1 and 2,
Table 2.1) designed for the linker. The sequencing was carried out at
Birmingham University ‘Alta Biosciences’. The forward primer (1) sequenced
the whole of the nos terminator and left border. The reverse primer (2)
sequenced through 700bp of the 870bp CaMV 35S promoter. The sequencing
confirmed the presence of the TATA box in the CaMV 35S promoter and the
polyadenylation signal (AATAA) in the nos terminator. The ACC cDNAs aco1,
aco2 and acs were sequenced with the cloning primers (5, 6, 7, 8, 9 and 10).
The insert sequences were aligned with the database sequences for the
broccoli aco1, aco2 and acs cDNAs by www.ncbi.nlm.nih.gov/BLAST/,
confirming the results.
61
2.4. Discussion
The Brassica oleracea L. var. italica ACC oxidases 1 and 2 and ACC
synthase cDNAs have been cloned into the minimal T-DNA vector pSCV1.0
between a CaMV 35S promoter and a nos terminator. To achieve this, a
cloning strategy with a few steps that could be efficiently executed was
devised. A backbone vector pJ1.0, was first constructed which could be used
universally to clone in independent ACC cDNAs. As there were limitations
with restriction sites, the - glucuronidase (gus) gene from pBI121 was
removed with XbaI and SstI and replaced by a synthetic linker. The synthetic
linker contained unique restriction sites in pSCV1.0 that could be used for
cloning. The CaMV 35S promoter-linker-nos terminator cassette was
transferred to the pSCV1.0 multiple cloning site as a HindIII/EcoRI inverted
fragment relative to the left and right borders.
Originally, the ACC oxidase 1 and 2 and ACC synthase cDNAs had
been ligated into the EcoRI site of pBluescript SK- after attachment of NotI
adaptors. The aco1 cDNA had been ligated in sense orientation, whereas the
aco2 and acs were in antisense orientation. There was a NotI restriction site
in pSCV1.0 so it was not possible to use these adaptors for cloning. The
cloning steps were also limited by restriction sites within the cDNAs. To
circumvent these problems, the cDNAs were amplified by polymerase chain
reaction (PCR) that provided ample quantites of DNA to complete the cloning
steps and avoid the need for doing minipreps and digests that are labour
intensive.
One of the objectives was to provide enough choice of restriction sites
to facilitate the use of alternative promoters. Blunt-ended ligation into the
62
SmaI site of pJ1.0 would leave HindIII or PstI with BamHI for aco1 and aco2
and BclI with BamHI for acs. The efficiency of blunt-ended cloning for aco1
and aco2 was about one colony in twenty with the insert. Enough colonies
were found using the colony PCR method. Blunt-ended cloning did not work
for ACC synthase (results not shown). ACC synthase is 1734bp, whereas
aco1 and aco2 are 1237bp and 1233bp, respectively. The extra 500bp made
this ligation too inefficient to proceed. The next progression in this strategy
was to put BamHI or SalI sticky-ends onto the cDNA by adding extra bases to
the cloning primers. The restriction enzyme SalI was not very efficient at
cutting pJ1.0. This meant that there were too many colonies which did not
have the insert and impossible to find those with an insert. Instead, BamHI
ends were added to the cDNA by PCR and ligated to the BamHI site of pJ1.0
at an efficiency of about 2 in 20 colonies. A consequence of this strategy was
that unlike ACC oxidase 1 and 2, the promoter cannot be changed. However,
it was successful in providing sense and antisense ACC synthase constructs.
In conclusion, an intermediate pJ1.0 backbone plasmid with the CaMV 35S promoter and nos terminator flanking a multiple cloning site has been
successfully constructed and used for the cloning of ACC oxidase 1 and 2 in
sense and antisense orientations, ACC synthase in sense and antisense
orientations, and a gus control. These constructs were introduced into
Agrobacterium rhizogenes strain LBA 9402 for the transformation of Brassica
oleracea L. var. italica to down-regulate genes involved in post-harvest
ethylene production and extend the shelf-life of this floral vegetable.
63
3.0 Transformation of GDDH33
with Agrobacterium rhizogenes
64
3.1 Introduction
The soil bacteria Agrobacterium rhizogenes and A. tumefaciens are
responsible for the development of hairy root and crown gall disease, respectively
(Gelvin, 1990). They are able to insert DNA sequences (T-DNA or transferred
DNA) via a natural system of genetic transformation, into the genome of mono-
and di-cotyledonous plant cells. The determinants of both hairy root and crown
gall disease are extra-chromosomal plasmids termed Ri (root inducing) and Ti
(tumour inducing) harboured by the virulent bacteria (Gelvin, 1990). Ri- and Ti-
plasmids are large (200 to greater than 800kbp) and contain two regions for
tumourigenesis, the T-DNA and the vir (virulence) regions (Gelvin, 1990). The vir
region contains genes involved in processing and transferring the T-DNA into the
plant cell, and the T-DNA carries genes that induce neoplastic plant growth when
expressed within the plant cell nucleus (Gelvin, 1990).
The natural capacity of Agrobacterium to genetically transform plant cells
has been utilised to produce genetically modified plants in a range of species
(Christou 1996). In brassica, A. rhizogenes and A. tumefaciens are widely used for
plant transformation, both having advantages and limitations (reviewed by
Poulsen, 1996 and Puddephat et al., 1996). Transformation of Brassica oleracea
L. var. italica with A. rhizogenes was easily distinguishable by the emergence of
hairy roots (Tempe and Casse-Delbart, 1987). These hairy roots were clonal,
arising from single cells (Tempe and Casse-Delbart, 1987), whereas Berthomieu
et al. (1994) with A. tumefaciens produced a high frequency of undesirable
chimeric plants. A. tumefaciens is often preferred by researchers (see Puddephat
et al. 1996 for review) as it does not contain the root inducing loci (rol genes) on
the Ri TL-DNA of A. rhizogenes that causes the hairy root syndrome in plants. This
65
phenotype typically exhibits wrinkled leaves, shortened internode lengths, non-
geotropic roots, reduced apical dominance, altered flower morphology, and
reduced seed production (Tepfer, 1990).
Transformation of Brassica oleracea with Agrobacterium rhizogenes and
tumefaciens is achieved by inoculating explants in tissue culture (Poulsen, 1996).
The inoculated cut plant surface stimulates the proliferation of untransformed
adventitious roots, and these are phenotypically similar to transgenic ‘hairy-roots’.
It is important that the transgenic roots are distinguished from the adventitious
roots, if they are to be isolated. Distinguishing between adventitious and A.
rhizogenes transformed roots has been limited by poor selection efficiency
achieved with antibiotic resistance genes (Hosoki and Kigo, 1994) whose
presence also leads to abnormalities that are regarded as disadvantageous (Earle
et al., 1996). A standard approach is to use reporter genes that allow the visual
detection of transgene expression without any need for a selectable marker. The
E. coli derived -glucuronidase gene (gus) (Jefferson et al., 1987) is one of the
most extensively used visual markers. It hydrolyses a wide variety of -
glucuronides that release bromo-chloro-indolyl compounds, staining the tissue
blue. GUS is very stable, but the histochemical assay is destructive. The firefly
luciferase enzyme, encoded by the luc gene, can be monitored in vitro (Ow et al.,
1986) but requires an exogenous substrate (luciferin) and emits light only at very
low intensity (Ow et al., 1986). The green fluorescent protein (GFP) from the
jellyfish Aequorea victoria requires no exogenous substrate and emits visible light
when excited by blue or UV light (Prasher, 1995). The use of the green fluorescent
protein to detect transformed tissue has been reported (Molinier et al., 2000;
Blumenhal et al., 1999 and Sheen et al., 1995) to be a rapid, non-invasive
66
technique. The gfp gene has been incorporated into the TL-DNA of the virulence
plasmid of A. rhizogenes strain LBA 9402 to produce the co-integrate vector
pRi1855::GFP (see Figure 3.1).
The co-integrate strain possesses the Ri TL-DNA encoding genes
necessary to produce transgenic hairy roots and the binary vector T-DNA carrying
the genes of interest to be co-transferred during the transformation process. Co-
transformation involves the stable integration of two species of T-DNA into the
plant genome, achieved with high efficiency (>50%) by Christey and Sinclair
(1992); Hosoki and Kigo (1994) and Puddephat et al. (2001) with vegetable
brassica plants. Co-transformation efficiency has proven to be greater with co-
integrate vectors (Fry et al., 1987; Charest et al., 1988) compared to binary
vectors when the T-DNAs are located on separate plasmids. If the two T-DNA
sequences integrate into the plant genome at physically unlinked loci it is possible
to segregate the Ri TL-DNA from the gene of interest T-DNA. The Ri TL-DNA
causes the ‘hairy root’ syndrome in plants, so it is important that it is segregated
out. This may be achieved by either sexual crossing, self-pollinating or through
microspore culture. Segregation frequencies of greater than 50% with one strain
and two plasmids have been demonstrated by Daley et al. (1998) and Komari et
al. (1996), with an A. tumefaciens protocol. More recently Puddephat et al. (2001)
has produced transgenic marker-free brassica plants with an A. rhizogenes
protocol.
The Brassica oleracea L. var. italica ACC oxidases 1 and 2 and ACC
synthase cDNAs (Pogson et al., 1995) and gus gene (Jefferson et al., 1987) have
been cloned into the minimal T-DNA vector pSCV1.0 (Biogemma) between a
cauliflower mosaic virus (CaMV) 35S promoter and a nopaline synthase (nos)
67
terminator (chapter 2). They have been transferred into the A. rhizogenes co- integrate strain LBA 9402 pRi1855::GFP by electroporation (chapter 2).
CaMV 35S gfp nos ter
3000 6000 9000 12000 18000 21000
3 6 8 10 12 13 14 R
1 2 4 5 7 9 11 15 16 17 18
rol A B C D
Figure 3.1 The TL-DNA of the agropine Ri plasmid pRi1855::GFP. It represents the relative location of each open reading frame (1-18) (Slightom et al., 1985) and the insertion of the CaMV 35S-gfp-nos construct by homologous recombination (Ian Puddephat pers. comm. HRI-Wellesbourne). Recombination was achieved by amplifying E36 (a sequence between 13 and 14) by PCR and ligating into pBR322 with the pBIN m-gfp5-ER construct. Homology between E36 and the virulent plasmid facilitated recombination. The polarity of each ORF is given by the position of the solid boxes: boxes above the line indicate transcripts running left to right, and boxes below the line indicate transcripts running right to left.
GDDH33 is a doubled-haploid line derived from the calabrese cultivar
Green Duke through anther culture. It has a homozygous genetic background, and
is amenable to transformation and regeneration through Agrobacterium
rhizogenes (H. Robinson, pers. comm. HRI-Wellesbourne) and is responsive to
microspore culture (L. Harvey, pers. comm. HRI-Wellesbourne). It is therefore a
suitable genotype for transformation and regeneration in these experiments.
68
The aim of this part of the project was to transform GDDH33 with A.
rhizogenes co-integrate strain LBA 9402 pRi1855::GFP harbouring the binary
vectors carrying the ACC constructs and gus reporter gene (Jefferson et al., 1987)
on the binary vector T-DNA. The target was to produce 20 independently
transformed root clones for each construct and then if possible, to regenerate
transgenic plants through tissue culture that could be tested for post-harvest
production of ethylene, and chlorophyll levels. If the T-DNAs from the
transformation events were physically unlinked, it would be possible to segregate
the two sequences when the T0 plants were self-pollinated to produce marker-free
plants.
69
3.2 Materials and Methods 3.2.1 Plant material and culture conditions
Seeds from the Brassica oleracea L. var. italica doubled-haploid line GDDH33
were used in transformation experiments. They were surface sterilised by immersion
in 1.7 % (w/v) sodium dichloroisocyanurate (BDH, UK) for 6 minutes followed by two
rinses in sterile (purified) water. Seeds were then germinated for three days on moist
filter paper in 9 cm Petri-dishes incubated at 22oC with 16 h light with a mix of 70 W
white and 65/80 W gro-lux fluorescent tubes providing an irradiance of 80 mol m-²
s-¹ at the culture level.
3.2.2 Plant transformation
The Agrobacterium rhizogenes co-integrate strain LBA 9402 pRi1855::GFP
(Spano et al., 1982, Puddephat et al., 2001) was used. The root inducing plasmid
pRi1855 carries the gfp reporter gene driven by the CaMV 35S promoter (see Figure
3.1). Seven binary plasmids were introduced into separate bacterial clones by
electroporation (Mattanovich et al., 1989). These binary plasmids were constructed
from pSCV1.0, a minimal T-DNA vector (Biogemma) with the CaMV 35S promoter
(Guilley et al., 1982) and nos terminator derived from pBI121 (Bevan, 1984;
Jefferson et al., 1987) (chapter 2). cDNAs of aminocyclopropane carboxylic acid
synthase and oxidase 1 and 2, isolated from Brassica oleracea by Pogson et al.
(1995) had previously been ligated (chapter 2) between the promoter and terminator
in both sense and antisense orientations, in relation to the promoter. The -
glucuronidase gene (gus) from Jefferson et al. (1987) had also been ligated into the
pSCV1.0 T-DNA region. Prior to plant transformation experiments, LBA 9402 was
sub-cultured on semi-solid YMB medium (see appendix). It was supplemented with
70
50 mg/l tetracyclin, 100 mg/l rifampicin and 50 mg/l kanamycin for selection of the
cointegrate Ri plasmid, chromosomal DNA and binary plasmid, respectively and
incubated at 25 C overnight.
From the overnight culture, three-four 10 l loops of Agrobacterium were used to inoculate 10 ml of MGL broth (see appendix). Broths were incubated for 16
hours at 25 C on a shaking platform (135-180g). Agrobacterium cells were
pelleted by centrifugation at 11600g for 5 minutes. Cells were resuspended in
liquid MS30 (see appendix) and supplemented with 1 mg/l 2,4-D to produce an
optical density (A600nm) of 1.0 (+/- 0.1).
Explants for use in transformation experiments were excised from six-day- old seedlings, grown under aseptic conditions, by cutting the hypocotyl
approximately 5 mm below the cotyledonary petioles. Explants were inverted and
placed on 6ml MS30 medium supplemented with 200mg/l cefotaxime in 5cm deep-
form Petri-dishes, with two explants per dish. A 2 l drop of re-suspended
Agrobacterium was placed on the cut surface of the hypocotyl.
The general components of the experimental set-up involved preparing the
Agrobacterium, excising the explants, inoculating the cut surface and labelling the
Petri-dishes. This order varied in two different experimental approaches. In the first
set of experiments (16), all of the explants were excised and then placed onto the
medium before being inoculated. In the later experiments (5), batches of 50
explants were excised, placed onto the medium and then inoculated. Inoculated
explants were incubated as previously described for germinating seedlings.
71
3.2.3 Identification of transformed ‘hairy roots’
Transgenic root production was determined 21-35 days after
inoculation. Explants were illuminated under long wave UV radiation (UVP
BLAK-RAY lamp, model B-100 AP) to detect GFP fluorescence in roots. The
numbers of explants producing excisable GFP roots (>1cm from root tip) were
counted. For selected GFP fluorescent roots, root tip sections were excised
and cultured on MS30 supplemented with 0.2mg/l naphthalene acetic acid
(NAA) and 200mg/l cefotaxime to establish transgenic root clones. Root
clones were transferred to fresh culture medium at 3-5 week intervals and
cultured at 25 C in the dark.
3.2.4 GUS detection
Histochemical assays for the detection of -glucuronidase activity were
performed using buffers from Jefferson et al. (1987) (see appendix) on 2cm
excised root tip sections.
3.2.5 PCR
A PCR protocol was designed to confirm the presence of T-DNAs in
root clones. Under aseptic conditions, a root section (1cm from the growing
tip) was excised by using fine forceps and placed onto a layer of Nescofilm
(Osaka, Japan). A sterile cocktail stick was used to remove just the growing
tip and 5mm root, which was placed into a 0.5ml microfuge tube with 50 l
water. The tube was then heated at 99 C for 3 minutes, placed on ice for two
minutes and then microfuged for 30 seconds at 11600g. Five microlitres of the
supernatant was taken for each PCR.
72
Primers were designed (see Table 3.1) to produce fragments from:
promoter/terminator of the aco and acs cDNAs, and internal primers on the
gus gene to confirm insertion of the genes of interest in plant tissue; the gfp
gene to detect the Ri plasmid; and the Virulence (vir) region on the Ri plasmid
which is not transferred, to determine whether amplification is due to
Agrobacterium contamination in the sample. The aco1 and aco2 primers were
designed to amplify the 250bp 3’ untranslated regions of the cDNA that share
only 44% nucleotide identity, to be used as probes. The acs F and R primers
amplify the whole cDNA region of ACC synthase to be used as a probe.
Table 3.1 Oligonucleotide sequences used as primers for the verification of T- DNA insertion and making DNA probes for hybridisation.
Names
Primer sequence (5`- 3`)*
Tm
( C)(a)
3 ACG TTC CAA CCA CGT CTT CA 60 4 AGC TTC ATC GCT TGC CTG TA 60 5 CTT CTC GAC GTT GTC CAT CA 60 8 CCA CAG TCG AGA CGT TCT AA 60 9 TTC ATG GCC GCT CGG TAT TC 62 12 AGA AGT GTT GGC AGA GTA GC 60 13 CTT CGC TAC AGC TTG TCT GA 60 nos R CCG ATC TAG TAA CAT AGA TGA CAC 61 vir F ATG TCG CAA GGA CGT AAG CCG A 65 vir R GGA GTC TTT CAG CAT GGA GCA A 63 gus F GCG TGG TGA TGT GGA GTA TT 60 gus R ACG CGG TGA TAC ATA TCC AG 60 gfp F GGC CAA CAG TTG TCA CTA CT 60 gfp R AAG AAG GAC CAT GTG GTC TC 60 aco1 3’UTR F GAA GAA TTC TAA TGC AGT TAC AG 57 aco1 3’UTR R CGC AAG CTT AGA ACA GTT ATC ATA 59 aco2 3’UTR F GAA GAA TTC TGC AGC AAC CAC GGA TTT G 66 aco2 3’UTR R GCG GAA GCT TTA AAA TTC ATA TTT CCA ATA C 62 acs cDNA F AAT AAC GAT TAG AGA AAT TAT GGT 54 acs cDNA R ATC TCA ACA GAA CAA AAA CAA AC 56
*Primers designed in relation to the inserts being in sense orientation to the promoter. (a) Tm of the primers was calculated with the expression:
Tm = 81.5 + -16.6 + (41 x (#G + #C/length)) -(500/length)
73
Table 3.2 PCR reaction components. To make a total 25 l for each reaction, 5 l of template DNA (100pmol/ l) and 5 l of primers (set at 4pmol/ l) were added. Primers and template DNA were suspended in autoclaved R.O water.
x1 ( l) 10x Buffer 2.5 dNTPs (25mM) 0.2 H2O 12.37 BRL Taq 0.13 Total 25
Table 3.3 PCR program The annealing temperature was adjusted to 3oC below the Tm of the primer
pairs. This was performed on a Hybaid Omnigene PCR machine. Stage 1 and 3 were cycled once. Stage 2 was cycled thirty five times.
Stage 1. 1) 93 C/ 2 min 2) 53 C/ 1 min
Stage 2. 1) 72 C/ 1 min 2) 93 C/ 30 sec 3) 53 C/ 1 min
Stage 3. 1) 72 C/ 10min.
All oligonucleotides were synthesised by MWG-Biotech (MWG-BIOTECH
AG Anzinger Str.7 D-85560 Ebersberg) (Table 3.1) and used in standard PCR
conditions (Tables 3.2 and 3.3). PCR components including Taq polymerase, PCR
buffer and dNTPs were obtained from GIBCO-Life Technologies, Uxbridge, UK.
74
3.2.6 Shoot Regeneration from established root clones selected by GFP fluorescence
Six root explants with 15-20mm sections were excised from
established root clones and cultured on MS30 medium, supplemented with
5mg/l NAA and 5mg/l benzylaminopurine (BA), and solidified with 6g/l agar in 5cm Petri-dishes. Cultures were incubated as previously described and root
explants producing callus transferred to fresh regeneration medium at three-
week intervals. Calli forming shoot initials were transferred to Gamborg’s B5
medium (see appendix) with 20g/l sucrose supplemented with 0.3mg/l Indole-
3-propionic acid (IPA) for shoot elongation. Single differentiated shoots with
roots were transferred to 40ml of regulator-free, MS30 medium in magenta
tubs (Sigma-Aldrich, Poole, Dorset, UK). All cultures were incubated as
previously described, for 3 weeks. The rooted shoots were transferred to non-
sterile compost (a 80:20 mix of Levington M2 and vermiculite) in 5cm
modules. They were maintained in humidified plant propagators up to 7 days,
progressively reducing the humidity before being transferred to 15cm pots.
They were grown in glasshouse conditions set to maintain 22 C air
conditioning with daylight extension to a minimum 14 hours photoperiod
provided by high pressure sodium lamps.
3.2.7 DNA extraction
Two grams (fresh weight) of the youngest leaves from post-headed
transgenic and non-transgenic broccoli was immediately frozen in liquid
nitrogen, before being freeze-dried at -80oC. The freeze-dried leaf material
was ground into a fine powder with a mortar and pestle, and 0.5g was
75
transferred to a 50ml polypropylene centrifuge tube. To this 15ml Kirby mix
(1% w/v sodium tri-isopropylnaphthalene sulphonate; 6% w/v sodium 4-amino
salicylate; 2.5%v/v 2M Tris-HCl pH8 and 13.5ml R.O H2O) was mixed and
shaken gently for 30min. 10ml phenol/chloroform was then added and shaken
gently for 5min, before being centrifuged at 2677g for 10min at room
temperature. The upper phase was removed and placed into a fresh 50ml
tube with 0.1 volumes of 3M sodium acetate pH 6.0 and 0.6 volumes of
propan-2-ol, and left for an hour. The tubes were then centrifuged at 2677g for
10min. Following this, the supernatant was poured off and the pellets were
allowed to air dry for 30min. The pellets were resuspended in 2ml TE and
40 g/ml RNase and incubated at 37oC for 1 hour. They were then transferred to a 15ml polypropylene tube and mixed with 2ml phenol/chloroform for 5min.
Following this, the tubes were centrifuged at 2677g for 10min. The upper
phases were then transferred to fresh tubes and precipitated with 0.1 volumes
of 3M sodium acetate pH 6.0 and 0.6 volumes of propan-2-ol, and left
overnight. After precipitation, the tubes were centrifuged at 2677g for 10min,
the supernatant poured off and pellet left to air dry for 1 hour. The pellets were
resuspended in 120 l TE and left at room temperature for 1 hour.
3.2.8 Endonuclease restriction digest of genomic DNA
The genomic DNA was quantified on a 1% (w/v) agarose gel and
ranged from 200ng/ l in the non-transformed control to 20ng/ l for 5/00 GUS
1. The restriction endonuclease SstI was chosen as the constructs contain a single site between the cDNA and nos terminator. The digest comprised of
40 l genomic DNA; 20 l, reaction 2 buffer; 5 l SstI enzyme; 10 l (10mg/ml)
76
spermidine; and 125 l H2O to make a final volume of 200 l. The reaction was
left overnight at 37oC. The digestion components were then mixed with 200 l
phenol, centrifuged at 11600g for 15sec and the aqueous phase removed. To
this, 100 l phenol and 100 l chloroform were mixed and centrifuged as
above. The aqueous phase was removed and mixed with 200 l phenol,
centrifuged as above and the aqueous phase removed. This was placed in a
0.5ml microfuge tube with 600 l 100% ethanol and 100mM sodium acetate.
The tubes were left for 1 hour, then centrifuged at 11600g for 30min. The
pellet was air dried for 5 min and resuspended in 15 l H2O.
3.2.9 Southern Blot
Two 0.8% (w/v) agarose gels were poured and then loaded with the
digestion products. The first gel was loaded in lanes 1-14 with: HindIII lambda
ladder (GIBCO-Life Technologies); non-transformed GDDH33; 28/00 aco1A
3; 34/00 aco1A 1; 34/00 aco1A 4; 32/00 aco1S 3; 33/00 aco1S 4; 3/00 aco2A 2; 33/00 aco2A 3; 5/00 aco2 S; 26/00 aco2 S 1; 20ng pBIN m-gfp5-ER BclI
digest; 20ng aco2 construct BclI digest; aco1 construct BclI digest. The
second gel was loaded in lanes 1-13 with: HindIII lambda ladder (GIBCO-Life
Technologies); non-transformed GDDH33; 7/00 acsA 12; 7/00 acsA 13; 9/00
acsA 1; 9/00 acsA 2; 7/00 acsS 7; 5/00 gus 1; 33/00 gus 3; 34/00 gus 2; 20ng
pBIN m-gfp5-ER BclI digest; 20ng gus construct BclI digest; 20ng acs
construct BclI digest. The gels were ran at 50 volts for 18h. They were stained
in 0.33 g/ l ethidium bromide, photographed, and then de-stained in R.O
water.
77
The blotting was performed by alkaline transfer onto Hybond N+ filters (Amersham) according to Sambrook et al. (1989).
3.2.10 Hybridisation between N+ filters and 32P dCTP labelled DNA probes
The Hybond N+ filters were placed onto nylon mesh that had been
wetted in 2x SSC (saline sodium citrate) wrapped tightly and placed in a
Hybaid hybridisation bottle. Fifty millilitres of hybridisation buffer (0.5M Na
phosphate pH 7.2; 70g/l sodium dodecyl sulphate [SDS]; 10mM EDTA, pH 8.0
and; 100 g/ml single stranded salmon sperm DNA) was added to the tubes at
65oC. The tubes were placed in a Hybaid rotisserie oven at 65oC and left to pre-hybridise for 3h. DNA probes were produced by PCR for regions of the
gus, gfp, aco1, aco2 and acs genes with primers shown in Table 3.1. The
DNA probes were labelled with 32P dCTP according to the Rediprime II
random prime labelling system kit (Amersham Pharmacia Biotech, Bucks.
UK). The labelled DNA was separated from the other components by a
sepharose column and counted with a scintillation counter. The 50mls of
hybridisation buffer were replaced with 10mls fresh buffer, and 2 x 107 counts
per minute labelled probe and left overnight at 65oC. The blots were washed
at medium stringency with 2 x SSC and 1% SDS four times at 65oC. The
filters were wrapped in Saranwap (Dow Chemicals) and placed into x-ray
cassettes with XAR5 film (Kodak) under a safelight. The cassettes were
placed at –80oC for 2 nights, and then the film was developed.
78
3.2.11 Seeds
Seeds were obtained from T0 plants by Linda Doyle at Horticulture
Research International, Wellesbourne, containment glasshouses by selfing.
Pollinations were carried out by hand at the unopened bud stage after
isolation of influorescences in cellophane bags, and seeds were harvested at
maturity.
3.2.12 Statistical Analysis
Dr J. Lynn at Horticulture Research International, Wellesbourne,
carried out the binomial regression analysis for the transformation results.
Students T-Test was also used to compare sample means.
79
3.3 Results
3.3.1 Constructs
The aim was to produce 20 independently-transformed GDDH33 root
lines for each construct. The number of inoculations was therefore dependent
on transformation efficiency. In total, 6566 GDDH33 seeds were chitted of
which 4599 (72.1%) germinated and were inoculated with A. rhizogenes LBA
9402 pRi1855::GFP strain, harbouring the binary plasmids carrying the
constructs. From 4599 inoculated explants, 150 produced GFP-fluorescent
excisable roots, giving an overall transformation efficiency of 3.63% 0.56
(Table 3.4). Transformation efficiency of the ACC synthase antisense
construct was significantly greater t(df) = 3.64(40), p<0.001.
Table 3.4 Transformation efficiency of LBA 9402 pRi1855::GFP carrying binary plasmids with the B. oleracea var. italica ACC cDNA constructs on GDDH33.
Construct No. Inoculations gfp+ve Roots Efficiency
(%)
aco1A 679 18 2.65
aco1S 685 18 2.63
aco2A 826 25 3.03
aco2S 622 20 3.22
acsA 294 21 7.14
acsS 474 18 3. 80
gus 1019 30 2.94
Total 4599 150 3.63 0.56
aco1 = ACC oxidase 1; aco2 = ACC oxidase 2; acs = ACC synthase; gus = - glucuronidase. Suffix, A = antisense, S = sense
80
3.3.2 Inoculation treatment
The inoculation of explants with A. rhizogenes was divided into two
separate treatments. In the first treatment, all of the explants were excised
before inoculation and, in the second, batches of 50 explants were excised
before inoculation. The second treatment produced significantly more
explants with excisable GFP+ve roots than the first treatment t(df) = 6.33(39),
p<0.001. Individually, transformation efficiency was improved for all constructs
in treatment 2, except ACC synthase sense which was non-significantly
reduced t(df) = 0.07(39), p>0.001 (Figure 3.2). There was no treatment two for
ACC synthase antisense as enough roots had been obtained in treatment
one. The greatest difference occurred between treatments for the gus
construct where transformation efficiency increased from 1.14-11.70%. This
ten-fold increase in transformation efficiency in treatment two was not
observed with the ACC constructs. It is assumed that the ACC constructs
were having a negative impact on transformation efficiency. Thus,
manipulation of ACC oxidase and ethylene biosynthesis appeared to have an
influence on transformation efficiency.
81
Tran
sfor
mat
ion
effic
ienc
y (%
)
16
14
12
10
8
6
4
2
0 ACO1A ACO1S ACO2A ACO2S ACSA ACSS GUS
Construct
Figure 3.2 The effect of treatment on construct transformation efficiency. The transformation efficiency of treatment 1 is shown by the blue bars and treatment 2, by the maroon bars. aco1 = ACC oxidase 1; aco2 = ACC oxidase 2; acs = ACC synthase; gus = -glucuronidase. Suffix, A = antisense, S = sense
82
3.3.3 gfp reporter gene
The green fluorescent protein (GFP) was a very effective reporter of
transgene expression in roots. Strong visual fluorescence was achieved
simply by holding the root sample under the long-wave UV lamp (Figure 3.3)
for 1-2 seconds. It was easy to distinguish between the bright ‘lime-green
fluorescing’ GFP-expressing roots and the non-fluorescing roots that had not
been transformed. This facilitated the removal and culture of only transgenic
roots.
3.3.4 Co-transformation
The efficiency of co-integral insertion of the root-inducing TL-DNA and
the T-DNA from the binary vectors was assessed using the gus construct.
Transgenic hairy roots expressing the gfp reporter gene inoculated with A.
rhizogenes carrying the gus construct were analysed histochemically for -
glucuronidase activity (Figure 3.4). Twenty-three out of twenty-seven roots
tested showed GUS activity. All of these roots were tested by PCR for
confirmation of the T-DNA inserts. The PCR demonstrated that all roots
contained gfp inserts, and only one had no evidence for gus gene insertion.
Therefore, only one from twenty-seven proved negative for the gus gene, and
positive for gfp resulting in a 96% co-transformation efficiency.
83
Figure 3.3 Nine-week-old GFP fluorescing root/callus. The tissue was grown on MS30 supplemented with 5mg/l NAA and 5mg/l benzylaminopurine (BA), and solidified with 6g/l agar in 5cm Petri-dishes.
Figure 3.4 Six-week-old GUS positive co-transformed root. The root was taken from the transformation event 32/00 6 (6th root from 32nd
experiment in year 2000). It was grown on MS30 supplemented with 0.2mg/l naphthalene acetic acid (NAA) and 200mg/l cefotaxime. The 2cm excised root was incubated with buffers from Jefferson et al. (1987).
84
3.3.5 Regeneration
All lines producing shoots in tissue culture regenerated at least one
whole plant in the glasshouse and the 18 regenerated lines produced a total
number of 195 plants (Table 3.5). The sense constructs produced fewer lines
(5) than the antisense (10), which was most clearly observed with the ACC
synthase sense (1) and antisense (4). The regeneration of shoots from
transgenic roots was a major limiting factor for producing transgenic plants.
The overall efficiency of explants inoculated to the production of transgenic
plants shows a mean of one explant in 200 (Table 3.5 and see Table 3.4 for
transformations), outlining the difficulty of the whole tissue culture process.
Table 3.5 Transgenic lines regenerated into whole plants
Construct No. lines producing
shoots
Efficiency of root
lines producing shoots (%)
Efficiency of lines producing shoots
from inoculations (%)
aco1A 3 16.67 0.44
aco1S 2 11.11 0.29
aco2A 3 12 0.36
aco2S 2 10 0.32
acsA 4 19.05 1.36
acsS 1 5.56 0.21
gus 3 10 0.29
Total 18 12.05 1.70 0.47 0.15
aco1 = ACC oxidase 1; aco2 = ACC oxidase 2; acs = ACC synthase; gus = - glucuronidase. Suffix, A = antisense, S = sense
85
(a) (b)
(c) (d)
Figure 3.5 The morphogenesis from root callus to plantlet. (a) Root callus; (b) Root callus producing shoots; (c) Single shoots from root callus (d) Small plantlet. Tissue in (a)-(c) is growing in 5cm Petri-dishes, and (d) in a magenta tub.
Root sections were placed onto a regeneration medium leading to a visible
swelling and yellow colouration after 3 weeks. Root production was generally
inhibited on this medium although it was not uncommon to have a root/callus
tissue as shown in Figures 3.3 and 3.5(b). Tissue morphogenesis proceeded 6-12
weeks with further expansion and the production of chlorophyll pigments as shown
in Figure 3.5(a). All of the root lines progressed into this callus phase but most
deteriorated into brown, necrotic tissue after 12 weeks on the shoot inducing
86
medium. Only tissue from 18 out of 150 lines reached the next stage (Figure 3.5b)
where calli produced anthocyanin pigments and developed into small shoots.
Once the single shoots had been defined (3-6 weeks), they were transferred to a
shoot elongation medium containing IPA (0.3mg/l) to develop and produce roots
(Figure 3.5c). It was important to establish a root system on this medium, as not
all of the lines were able to produce roots on hormone-free medium. Once the root
systems had been established, the rooted shoots were transferred to hormone-
free MS30 (Figure 3.5d), and grown for 6-9 weeks within magenta tubs. They were
transplanted into trays containing compost in the glasshouse and moved into
larger pots to be grown into whole T0 plants.
The average time between the inoculation of the explant and the production of a mature broccoli head through regeneration was 389 16d, just under a year
and a month. The lines transformed with the gus construct matured in 317 1.5d,
significantly faster t(df) = 4.5(15), p<0.001 than those with the ACC constructs that
took 404 16d. Interestingly, the gus lines actually took significantly longer t(df) =
5.9(15), p<0.001 to mature in the glasshouse with 168 14d, than the other lines
with 106 8d. The overall difference arises because the ACC constructs had
significantly reduced the efficiency of regenerating a transgenic hairy root into a
plantlet, t(df) = 6.4(15), p<0.001. It had taken 149 55d from inoculating to
transferring the gus plantlets into the glasshouse, whereas it had taken 291 21d
for the other lines.
3.3.6 Determining stable insertion of the T-DNAs into the plant genome
DNA was extracted from leaf material of lines regenerated from roots
through tissue culture. Specific primers to the gfp gene were used to check for
87
the presence of the TL-DNA region of the A. rhizogenes virulence plasmid
pRi1855::GFP, and for the insertion of the binary plasmid T-DNA containing
the 7 constructs.
In Figure 3.6 (lanes 2-4), the lines inoculated with A. rhizogenes
containing the aco1 antisense constructs produced the 542bp band that would
be expected if the target DNA was present. There are smears in lanes 5 and
6, and it is difficult to determine a 360bp band for the presence of aco1 sense
constructs. The aco2 antisense lines in lanes 7 and 8 have produced the
expected band of 369bp as with the two aco2 sense lines in lanes 9 and 10
producing a 578bp band. Lanes 11-14 contain the lines derived from
transformation with the acs antisense constructs. However, only 7/00acsA 12
and 9/00 acsA 1 have produced the 471bp fragment expected from
amplification of the target DNA (transformation code= 7th experiment in the
year 2000, acs antisense construct, and 12th explant with gfp fluorescing excisable root). The 7/00 acsS 7 line in lane 15 has produced the expected
band of 530bp. Lanes 16-18 contain the 3 lines inoculated with the gus
construct. There is only a distinctive band of 390bp in lane 18, 34/00 gus 2.
This is the first time the PCR has not worked for all of the gus lines from this
set of DNA.
A PCR was carried out on regenerated lines for the presence of the gfp gene as shown in Figure 3.7. Although there is a smear, it is clear that the
494bp band of the gfp gene has been amplified in all of the lines except 7/00
acsS 7 where it is difficult to ascertain.
88
bp 1353 1078 872
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Figure 3.6 PCR of genomic DNA from the transgenic plants to confirm insertion of the T-DNA constructs from the binary plasmid. Lane1, X174RFDNA/HaeIII fragments; lane 2,28/00 aco1A 3 primers 3&4; lane 3, 34/00 aco1A 1 primers 3&4; lane 4, 34/00 aco1A 4 primers 3&4; lane 5, 32/00 aco1 S 2 primers 3&5; lane 6, 33/00 aco1S 4 primers 3&5; lane 7, 3/00 aco2A 2 primers 3&8; lane 8, 33/00 aco2 A 3 primers 3&8; lane 9, 5 aco2S 1 primers 3&9; lane 10, 26/00 aco2S 1 primers 3&9; lane 11, 7/00 acsA 12 primers 3&12; lane 12, 7/00 acsA 13 primers 3&12; lane 13, 9/00 acsA 1 primers 3&12; lane 14, 9/00 acsA 2 primers 3&12; lane 15, 7/00 acsS 7 primers 3&13; lane 16, 5/00 gus 1 primers gus F&R; lane 17, 33/00 gus 3 primers gus F&R; lane 18, 34/00 gus 2 primers gus F&R.
bp
1353 1078 872
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Figure 3.7 PCR of the gfp gene from genomic DNA of the transgenic lines to confirm insertion of the TL-DNA region of the virulence plasmid. All PCRs carried out with the GFP F&R primers. Lane1, X174RFDNA/ HaeIII
fragments; lane 2,28/00 aco1A 3; lane 3, 34/00 aco1A 1; lane 4, 34/00 aco1A 4; lane 5, 32/00 aco1 S 2; lane 6, 33/00 aco1S 4; lane 7, 3/00 aco2A 2; lane 8, 33/00 aco2 A 3; lane 9, 5 aco2S 1; lane 10, 26/00 aco2S 1; lane 11, 7/00 acsA 12; lane 12, 7/00 acsA 13; lane 13, 9/00 acsA 1; lane 14, 9/00 acsA 2; lane 15, 7/00 acsS 7; lane 16, 5/00 gus 1; lane 17, 33/00 gus 3; lane 18, 34/00 gus 2.
89
The PCR was a rapid tool to detect presence of the transgene but
could not determine stable insertion and copy numbers of particular
transgenes. The DNA extracted from the lines regenerated through tissue
culture and that from non-transformed GDDH33 was digested with the SstI
endonuclease. It was ran on a 0.8% (w/v) agarose gel at 15V for 18h as
shown in Figures 3.8 and 3.9 and blotted onto Hybond N+ filters (Amersham).
The filters were then hybridised to specific probes, and the results are shown
in Figures 3.10-3.15. Figures 3.8 and 3.9 show the quantity of DNA in each
lane of the transgenic lines digested by SstI as a reference for the
autoradiographs in Figures 3.10-3.15.
Figure 3.10 shows the autoradiograph of the hybridisation between the
Brassica oleracea var. italica aco1 250bp 3’UTR DNA probe and the DNA
shown in Figure 3.8. Lane 14 shows the production of a large signal where
there is hybridisation of the probe to the control DNA. There are no bands in
lane 2, as the specific probe has not strongly hybridised to the non-
transformed GDDH33 DNA even at medium stringency. The aco1 antisense
line in lane 3 has two distinct bands between the 6557 and 4361markers. At
the 2027 marker, there is a band with less intensity than the upper two. It
occupies the same position as one of the distinctive DNA bands on Figure
3.8, and may have produced a signal due to weak binding and a high copy
number. This lower band is possibly not the result of a T-DNA insert.
There are no clear distinctive bands in lanes 4-6. Lane 6 did not
actually contain the DNA of 32/00 aco1S 3 as it had diffused from the well just
after loading, as can be seen in Figure 3.8. This also occurred with 34/00
aco1A 1, although a small proportion remained within the well. There should
90
be enough DNA in lane 5 for 34/00 aco1A 4 for hybridisation with the aco1 3’UTR probe suggesting that this is a negative result. This is also possibly the
case with 33/00 aco1S 4 in lane 7, but there is a positive band at 6557bp but
due to the high background between the 23130 and 4361 markers is less
defined.
As expected, there are no bands in lanes 8-10 containing the aco2
lines 3/00 aco2A 2, 33/00 aco2A 3 and, 5/00 aco2S 1. This is not the case
with 26/00 aco2S 1 in lane 11 that contains 8 visible bands. They appear to
be genuine, but if compared to Figure 3.11, and the aco2 3’UTR probe, all of
these bands are in common. There is a higher concentration of DNA from
26/00 aco2S 1 compared to the other lines that might relatively amplify a
weak signal.
Figure 3.11 shows the autoradiograph of the hybridisation between the
Brassica oleracea var. italica aco2 283bp 3’UTR DNA probe and the DNA
shown in Figure 3.8. There are no bands in the non-transformed GDDH33
control DNA in lane 2. There are bands in lanes 5, 7, 8 and 11 at 2027bp
corresponding to the visible DNA bands in Figure 3.11. These also occurred
in Figure 3.10 and probably do not correspond to genuine aco2 T-DNA
inserts. There is a distinctive band in lane 7 between 23130 and 9416bp
positioned within the largest fragments of DNA, but possibly not a genuine
insert. There are no clear bands in lanes 9 and 10 containing the DNA of
33/00 aco2A 3 and 5/00 aco2S 1. In lane 11 with DNA from 26/00 aco2S 1,
there are 10 distinctive bands. Eight of these correspond directly to the bands
that hybridised to the aco1 3’UTR probe in Figure 3.10. The largest band, just
91
below the 6557bp marker and the fifth band down are not present in Figure 3.10 and probably represent the genuine bands for aco2 T-DNA insertion.
Figure 3.12 shows the autoradiograph of the hybridisation between the
Brassica oleracea var. italica acs full length cDNA probe and the DNA shown
in Figure 3.9. In Figure 3.12, there are bands in all of the lanes between the
23130 and 9416bp markers. As this band is present in the non-transformed
control it is probably not the result of a T-DNA insertion. There are 3
distinctive bands in lane 7 of 7/00 acsS 7 that are not present in the other
lanes. There is one band greater than 4361bp, one that size, and one
between 4361 and 2322bp that could be the result of 2-3 T-DNA insertions.
There is no evidence for stable insertion of acs antisense T-DNAs from this
experiment. The DNA from the SstI endonuclease restriction of 9/00 acsA 1
and 9/00 acsA 2 had diffused from the wells of lanes 5 and 6 and therefore
did not contain any DNA. There was DNA in lanes 3 and 4 for 7/00 acsA 12
and 7/00 acsA 13 that would be expected to have buds if the T-DNA insert
was present.
Figure 3.13 shows the autoradiograph of the hybridisation between the
gus 390bp DNA probe and the DNA shown in Figure 3.9. There are 9 strong
bands between 6557 and 2027bp in lane 10 with the DNA from 34/00 gus
2.The gus probe has not hybridised to the other transgenic lines or the non-
transformed control. It is possible that these are 9 genuine bands, although
one band between 4361 and 2322bp produces a much stronger signal and
may be the only genuine band. There was no 5/00 gus 1 DNA in lane 8 as it
had diffused from the well. There was however, DNA in lane 9, but no bands
could be a negative result.
92
Figure 3.14 shows the autoradiograph of the hybridisation between the gfp 494bp DNA probe and the DNA shown in Figure 3.8. In lanes 2-11 (except
6 as no DNA), there are bands at 2322bp at the same position as the
distinctive DNA bands in Figure 3.8. There is a band in the non-transformed
control (lane 2) as well as the lines regenerated through tissue culture. The
line 28/00 aco1A 3 in lane 3 has produced 4 distinctive bands between 23130
and 4361bp. In lane 4 with 34/00 aco1A 1, there is a band below 9416. There
are no clear bands in lanes 5-7 for lines 34/00 aco1A 4, 32/00 aco1S 3 and
33/00 aco1S 4. In lane 8 for 3/00 aco2A 2, there are clearly 3 bands between
the 9416 and 4361bp markers. There is not a band in lane 9, 33/00 aco2A 3,
but there is a singular band at 9416bp in lane 10 of 5/00 aco2S 1. In the
autoradiograph, there are clearly 5 distinctive bands in lane 11 for 26/00
aco2S 1, but is less obvious in the scanned image.
Figure 3.15 shows the autoradiograph of the hybridisation between the
gfp 494bp DNA probe and the DNA shown in Figure 3.9. There is a band in
the non-transformed control in lane 2 below the 4361bp marker consistent
with the DNA band in Figure 3.9. This band is also present in the lanes with
DNA from the regenerated lines. Lanes 3 and 4 containing DNA from 7/00
acsA 12 and 7/00 acsA 13 produced distinctive bands at 9416bp. There is no
DNA in lanes 5 and 6 due to diffusion from the well. In lane 7 there are 3
bands for 7/00 acsS 7 and 10 bands in lane 10 with 34/00 gus 2 although
there is one strong band between 9416 and 6557bp.
93
bp
23130
9416
6557
4361
2322
2027
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Figure 3.8 Genomic DNA from aco1 and aco2 transgenic lines digested with SstI restriction endonuclease and run on a 0.8% (w/v) agarose gel at 50V for 18h. Lane 1, Lambda HindIII fragments; lane 2, non-transformed GDDH33; lane 3, 28/00 aco1A 3; lane 4, 34/00 aco1A 1; lane 5, 34/00 aco1A 4; lane 6, 32/00 aco1 S 2; lane 7, 33/00 aco1S 4; lane 8, 3/00 aco2A 2; lane 9, 33/00 aco2 A 3; lane 10, 5 aco2S 1; lane 11, 26/00 aco2S 1; lane 12, pBIN m-gfp5-ER BclI digest; lane 13, aco2 construct BclI digest; lane 14, aco1 construct BclI digest.
94
bp
23130
9416
6557
4361
2322
2027
1 2 3 4 5 6 7 8 9 10 11 12 13
Figure 3.9 Genomic DNA from acs and GUS transgenic lines digested with SstI restriction endonuclease and run on a 0.8% (w/v) agarose gel at 50V for 18h. Lane 1, Lambda HindIII fragments; lane 2, non-transformed GDDH33; lane 3, 7/00 acsA 12; lane 4, 7/00 acsA 13; lane 5, 9/00 acsA 1; lane 6, 9/00 acsA 2; lane 7, 7/00 acsS 7; lane 8, 5/00 gus 1; lane 9, 33/00 gus 3; lane 10, 34/00 gus 2; lane 11, pBIN m-gfp5-ER BclI digest; lane 12, gus construct BclI digest; lane 13, acs construct BclI digest.
aco1 S 2; lane 7, 33/00 aco1S 4; lane 8, 3/00 aco2A 2; lane 9, 33/00 aco2 A 3; lane 10, 5 aco2S 1; lane 11, 26/00 aco2S 1; lane 12, pBIN m-gfp5-ER BclI digest; lane 13, aco2 construct BclI digest; lane 14, aco1 construct BclI digest.
95
bp
23130
9416
6557
4361
2322
2027
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Figure 3.10 Autoradiograph of the hybridisation between the Brassica oleracea L. var. italica aco1 283bp 3’UTR DNA probe and the digested genomic DNA of Figure 3.8. Lane 1, Lambda HindIII fragments; lane 2, non-transformed GDDH33; lane
3, 28/00 aco1A 3; lane 4, 34/00 aco1A 1; lane 5, 34/00 aco1A 4; lane 6, 32/00
aco1 S 2; lane 7, 33/00 aco1S 4; lane 8, 3/00 aco2A 2; lane 9, 33/00 aco2 A 3; lane 10, 5 aco2S 1; lane 11, 26/00 aco2S 1; lane 12, pBIN m-gfp5-ER BclI digest; lane 13, aco2 construct BclI digest; lane 14, aco1 construct BclI digest.
96
bp
23130
9416
6557
4361
2322
2027
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Figure 3.11 Autoradiograph of the hybridisation between the Brassica oleracea L. var. italica aco2 252bp 3’UTR DNA probe and the digested genomic DNA of Figure 3.8. Lane 1, Lambda HindIII fragments; lane 2, non-transformed GDDH33; lane 3,
28/00 aco1A 3; lane 4, 34/00 aco1A 1; lane 5, 34/00 aco1A 4; lane 6, 32/00
97
bp
23130
9416
6557
4361
2322
2027
1 2 3 4 5 6 7 8 9 10 11 12 13
Figure 3.12 Autoradiograph of the hybridisation between the Brassica oleracea L. var. italica acs full length cDNA probe and the digested genomic DNA of Figure 3.9. Lane 1, Lambda HindIII fragments; lane 2, non-transformed GDDH33; lane 3, 7/00 acsA 12; lane 4, 7/00 acsA 13; lane 5, 9/00 acsA 1; lane 6, 9/00 acsA 2; lane 7, 7/00 acsS 7; lane 8, 5/00 gus 1; lane 9, 33/00 gus 3; lane 10, 34/00 gus 2; lane 11, pBIN m-gfp5-ER BclI digest; lane 12, gus construct BclI digest; lane 13, acs construct BclI digest.
98
bp
23130
9416
6557
4361
2322
2027
1 2 3 4 5 6 7 8 9 10 11 12 13
Figure 3.13 Autoradiograph of the hybridisation between the 390bp gus DNA probe and the digested genomic DNA of Figure 3.9. Lane 1, Lambda HindIII fragments; lane 2, non-transformed GDDH33; lane 3, 7/00 acsA 12; lane 4, 7/00 acsA 13; lane 5, 9/00 acsA 1; lane 6, 9/00 acsA 2; lane 7, 7/00 acsS 7; lane 8, 5/00 gus 1; lane 9, 33/00 gus 3; lane 10, 34/00 gus 2; lane 11, pBIN m-gfp5-ER BclI digest; lane 12, gus construct BclI digest; lane 13, acs construct BclI digest.
99
bp
23130
9416
6557
4361
2322
2027
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Figure 3.14 Autoradiograph of the hybridisation between the 494bp GFP DNA probe and the digested genomic DNA of Figure 3.8. Lane 1, Lambda HindIII fragments; lane 2, non-transformed GDDH33; lane 3, 28/00 aco1A 3; lane 4, 34/00 aco1A 1; lane 5, 34/00 aco1A 4; lane 6, 32/00 aco1 S 2; lane 7, 33/00 aco1S 4; lane 8, 3/00 aco2A 2; lane 9, 33/00 aco2 A 3; lane 10, 5 aco2S 1; lane 11, 26/00 aco2S 1; lane 12, pBIN m-gfp5-ER BclI digest; lane 13, aco2 construct BclI digest; lane 14, aco1 construct BclI digest.
100
bp
23130
9416
6557
4361
2322
2027
1 2 3 4 5 6 7 8 9 10 11 12 13
Figure 3.15 Autoradiograph of the hybridisation between the 494bp GFP DNA probe and the digested genomic DNA of Figure 3.9. Lane 1, Lambda HindIII fragments; lane 2, non-transformed GDDH33; lane 3, 28/00 aco1A 3; lane 4, 34/00 aco1A 1; lane 5, 34/00 aco1A 4; lane 6, 32/00 aco1 S 2; lane 7, 33/00 aco1S 4; lane 8, 3/00 aco2A 2; lane 9, 33/00 aco2 A 3; lane 10, 5 aco2S 1; lane 11, 26/00 aco2S 1; lane 12, pBIN m-gfp5-ER BclI digest; lane 13, aco2 construct BclI digest; lane 14, aco1 construct BclI digest.
101
A summary of the data presented in Figures 3.6-3.15 is shown in Table
3.6. It is apparent that lines positive for construct and gfp by PCR are not
always confirmed by the Southern analysis. The copy number for construct
and that for the Ri-T-DNA insert ranges from 1-4 (or 11).
Table 3.6 Summary of the PCR and autoradiograph data for confirming the presence of T-DNAs in lines regenerated through tissue culture.
PCR Confirmation Southern Analysis (Copy No.)
Line Construct gfp Construct gfp
28/00 aco1A 3 2 4
34/00 aco1A 1 - -
34/00 aco1A 4 - -
32/00 aco1S 3 ? No DNA No DNA
33/00 aco1S 4 ? 1 -
3/00 aco2A 2 1 3
33/00 aco2A 3 - -
5/00 aco2S 1 - 1
26/00 aco2S 1 2-10 1-7
7/00 acsA 12 - 1
7/00 acsA 13 ? - 1
9/00 acsA 1 No DNA No DNA
9/00 acsA 2 ? No DNA No DNA
7/00 acsS 7 3 3
5/00 gus 1 - -
33/00 gus 3 - -
34/00 gus 2 1-10 1-11
102
3.3.7 Transgenic plant phentoype
The morphology of regenerated transgenic plant lines varied from
phenotypically normal to those with severe ‘hairy root’ syndrome. Four T0 lines
produced dwarfed plants (see Figure 3.16, a and c) with small, extremely
wrinkled leaves; shortened internodes; and small heads with reduced
chlorophyll. There were six T0 lines exhibiting moderate phenotypes with
wrinkled leaves and shortened internodes (Figure 3.16b), and eight (Figure
3.16c) that were phenotypically normal. It was possible to self-pollinate and
take buds for microspore culture from the moderate/normal plants, but not
from the plants with severe phenotype. Transgenic lines for these phenotypes
were scored as shown in Table 3.7.
As shown by Table 3.7, there is no apparent correlation between the
copy number of the Ri-T-DNA and the severity of the rol phenotype. There is
not enough data to determine whether there is a pattern for selection of Ri
phenotype within the constructs.
103
(a) (b)
(c) (d)
Figure 3.16 Heading broccoli plants that have been regenerated from GFP fluorescing roots, through a tissue culture phase. They illustrate different severities of the hairy root phenotype. (a) Severe rol phenotype (9/00 acsA 1; (b) Moderate rol phenotype (9/00 acsA 2); (c) Weak rol phenotype (7/00 acsA 12; (d) The most severe rol phenotype (5/00 aco2A 1).
104
Table 3.7 The effect of the rol genes on transgenic plant morphology.
Line gfp Copy No. rol phenotype
28/00 aco1A 3 4 weak
34/00 aco1A 1 NK weak
34/00 aco1A 4 NK weak
32/00 aco1S 3 NK weak
33/00 aco1S 4 NK moderate
3/00 aco2A 2 3 moderate
5/00 aco2A 1 NK severe
33/00 aco2A 3 NK weak
5/00 aco2S 1 1 moderate
26/00 aco2S 1 1-7 weak
7/00 acsA 12 1 weak
7/00 acsA 13 1 moderate
9/00 acsA 1 NK severe
9/00 acsA 2 NK moderate
7/00 acsS 7 3 severe
5/00 gus 1 NK severe
33/00 gus 3 NK weak
34/00 gus 2 1-11 moderate
NK=Not known
Plants transferred to the glasshouse from tissue culture were very
susceptible to abiotic stresses such as water loss leading to desiccation, and biotic
stresses through fungal pathogens and feeding insects. There was also a
detrimental effect on the plants through the action of the rol genes. In Table 3.8,
the contrast between the percentage of surviving plants with the normal phenotype
(89) and the severe phenotype (31) is indicative of the negative impact of the rol
genes.
105
Table 3.8 The total numbers of transgenic plants produced from all constructs with the effect of the rol genes on mortality.
rol phenotype No. plants transferred
to glasshouse
No. plants to surviving
heading
Percentage of plants surviving to heading
Weak 106 94 89
Moderate 86 58 67
Severe 70 22 31
All phenotypes 262 174 66
3.3.8 Seed production of the transgenic lines
The rol genes had a negative effect on seed production as shown in Table
3.9 as the lines with severe phenotypes did not produce any seeds. It also shows
that the constructs had little effect on as seed production. Seeds were produced
from lines with each construct except for ACC synthase sense, which only
produced one line that had a severe rol phenotype.
106
Table 3.9 The number of seeds recovered from transgenic lines with their particular rol phenotype
Plant Line No. of Seeds rol phenotype
03/00 aco2 A 2 39 moderate
07/00 acs A 12 454 weak
09/00 acs A 2 2 moderate
26/00 aco2 S 1 749 weak
28/00 aco1 A 3 564 weak
32/00 aco1 S 3 2048 weak
33/00 aco1 S 4 42 moderate
33/00 aco2 A 3 2075 weak
34/00 aco1 A 1 16 weak
34/00 aco1 A 4 20 weak
34/00 gus 2 13 moderate
Total lines = 11 Total seeds = 6022 4 = moderate
7 = weak
The transgenic line 28/00 aco1A 3 produced 564 seeds, of which 69
were immediately sown. Sixty-seven of these germinated (97%), and were
tested by PCR for presence of the gfp and aco1A T-DNA inserts. Figures 3.17
and 3.18 show the presence or absence of a band for the gfp and aco1A T-
DNA inserts with primers 3 and 4 (for aco1A) and gfp F and R (for gfp).
107
(a)
bp
1353 1078 872
(b)
bp 1353 1078 872
(c)
bp
1353 1078 872
(d)
bp
1353 1078 872
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Figure 3.17 Determination of gfp inserts by PCR of 67 28/00 aco1A 3 T1
progeny grown from seed. In lane 1 (a-d) X174RFDNA/HaeIII fragments. In (a) lane 2, gfp+ve control;
lanes 3-20 contain seedlings 1-18. In (b), lanes 2-20 contain seedlings 19-37. In (c), lanes 2-20 contain 38-56. In (d), lanes 2-11 contain seedlings 57-67.
108
(a)
bp 1353 1078 872
(b)
bp 1353 1078 872
(c)
bp
1353 1078 872
(d)
bp 1353 1078 872
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Figure 3.18 Determination of aco1A inserts by PCR of 67 28/00 aco1A 3 T1
progeny grown from seed. In lane 1 (a-d) X174RFDNA/HaeIII fragments. In (a) lane 2, aco1A+ve control; lanes 3-20 contain seedlings 1-18. In (b), lanes 2-20 contain seedlings 19-37. In (c), lanes 2-20 contain 38-56. In (d), lanes 2-11 contain seedlings 57-67.
109
The data in Figures 3.17 and 3.18 is summarised in Table 3.10. It
groups the seedlings into four categories depending on presence or absence
of the gfp and aco1A T-DNA inserts. There are a high proportion of double-
positives (57%) in Table 3.10 as the gfp and aco1A T-DNAs have co-
segregated. There are half those numbers of double-negatives where none of
the T-DNAs have been passed on. The smallest group with only 3% of the
progeny are the marker-free plants with only the construct T-DNA.
Table 3.10 Segregation of T-DNA inserts in the T(1) progeny of 28/00 aco1A 3
Segregation of T-DNA
inserts
No. of plants
gfp+ve/aco1A+ve 38
gfp+ve/aco1A-ve 10
gfp-ve/aco1A+ve 2
gfp-ve/aco1A-ve 17
Total 67
110
3.4 Discussion
The overall transformation efficiency achieved by inoculating GDDH33 with
the A. rhizogenes co-integrate strain LBA 9402 pRi1855::GFP was 3.63 0.60%.
This refers to the proportion of explants that produced excisable (>1cm) GFP-
fluorescing transgenic roots from those that were inoculated. This value is
comparable to broccoli transformation efficiencies of Green Comet 3.1, Corvet 4.2,
New River 6.7, and Vantage 2.1 achieved by Puddephat et al. (2001). It was not
as great as the 20% efficiency achieved by Cogan et al. (2001) for GDDH33,
although not all of the GFP-fluorescing roots would have been excisable by Cogan
et al. (2001). It is also possible that GFP expression was reduced in some of the
transgenic roots by co-suppression. Al-Kaff et al. (1998) silenced GUS expression
in transgenic Brassica napus carrying the gus gene and a CaMV 35S promoter by
infecting it with the CaMV. The homology of the promoter sequence was
suggested to lead to the post-transcriptional silencing of the gene. The ACC
constructs and the gfp gene both shared the CaMV 35S promoter, in the
transformation experiments.
Manipulating either the explant and/or the bacterium to enhance virulence
can increase transformation efficiency (Henzi et al., 2000). In this work, transgenic
root production was significantly improved by manipulation of the explant in
treatment 2. In treatment 2, explants were inoculated after the excision of only 50
seedlings, whereas in treatment 1 this number could be up to 208. The shorter
handling time in treatment 2, suggests that the newly excised tissue was more
susceptible to Agrobacterium-mediated transformation. Braun (1954, 1978) found
that tomato remained susceptible to tumourigenesis for up to two weeks after
excision but these plants even when treated with acetosyringone never achieved
111
susceptibility equal to plants inoculated directly after wounding. Inoculating plants
directly increased transformation efficiency in these experiments but it was still
very low. For many plant species or cultivars/ecotypes, T-DNA integration is the
most limiting factor for transformation protocols. This has been demonstrated by
relatively efficient transient expression of reporter genes but poor stable
transformation (Gelvin, 2000). It is therefore important to enhance the virulence of
the bacterium, susceptibility of the plant, and the rate of stable T-DNA integration
into the plant genome to optimise transformation efficiencies.
Transformation efficiency was observed to vary between the constructs
used. Inoculation of GDDH33 with the A. rhizogenes strain carrying the ACC
synthase antisense construct produced significantly greater numbers of explants
with excisable transgenic roots (p<0.001). Most plants species contain multiple
ACC synthase genes that can be quite divergent in DNA sequence, so it is
possible that the function of the ACC synthase identical to the construct may have
been silenced but replaced by another. Transformation efficiency increased 10-
fold, with the inoculation of the gus construct in treatment 2, but not with the ACC
constructs. The gus construct should not affect ethylene biosynthesis whereas it is
possible with the ACC constructs. This suggests that regulation of this plant
hormone is important for Agrobacterium-mediated transformation. Konings and
Jackson (1979) working with tomato, rice and white mustard found that roots
which elongated more rapidly, produced ethylene at the highest rate, and those
species whose roots grew slowly produced only small amounts of the hormone.
Chi et al. (1990) observed that silver nitrate (an inhibitor of ethylene perception)
inhibited rooting, and AVG (an inhibitor of ethylene biosynthesis) inhibited root
elongation in Brassica juncea plants. The gus control and the ACC synthase
112
antisense root lines produced rapidly growing roots. The constructs with poor
transformation efficiencies produced roots with poor growth in vitro that often led to
root mortality after a few subcultures.
Transgenic root growth was generally poor throughout all of the
transformed GDDH33 lines. They did not show the characteristic plagiotropic
growth as seen in other Brassica cultivars such as Shogun (Henzi et al., 2000).
Berthomieu and Jouanin (1992) similarly transformed a doubled haploid line of a
rapid cycling cabbage (Brasscia oleracea L. var. capitata) with A. rhizogenes. Nine
roots were cultured of which four grew fast in hormone-free medium, but after two
subcultures, significantly reduced their growth rate. The other five transformed
roots grew very poorly and could not be significantly improved by using media with
additional macro-elements or hormones. The lack of root growth is probably
genotype-dependent, as other cultivars have proved very amenable to rapid
growth in tissue culture (Henzi et al., 2000).
The green fluorescent protein (GFP) was an excellent in planta reporter of
transgene expression in Brassica hairy roots. It was a rapid, non-destructive
technique that did not require exogenous substrates, thus making it more effective
than the gus reporter gene (Jefferson et al., 1987) and the luc gene (Ow et al.,
1986). Hairy roots expressing GFP were clearly distinguishable from those that
had not been transformed, and were easily removed and cultured.
Inoculation with the A. rhizogenes co-integrate strain produced 96% co-
transformed roots. This was higher than the 50% achieved by Komari et al. (1996)
when rice and tobacco were co-cultivated with an Agrobacterium strain carrying a
super binary plasmid with two different T-DNAs, and greater than that of Depicker
et al. (1985) who reported co-transformation frequencies of 60-70% with tobacco
113
protoplasts. The success of co-transformation may be attributed to the co-integrate
strain. It minimised the types of DNA being transferred to two: the root inducing TL-
DNA and the gene of interest T-DNA, and the need for only one strain. Gelvin
(1990) demonstrated the high efficiency of T-DNA transference by Agrobacterium
through transient gene expression, so the limiting factor may have been Ri TL-DNA
integration. When the 21Kbp Ri TL-DNA was integrated into the plant genome,
96% of the time the smaller 2-3Kbp binary vector T-DNA was also integrated. McCormac et al., (2001) achieved 100% co-transformation of Nicotiana tabacum
with the premise that the co-transformed T-DNA must be 2-fold smaller than the
selected T-DNA region. If the two T-DNA sequences have integrated into the plant
genome at physically unlinked sites it provides the opportunity to recover marker-
free, phenotypically normal plants as shown by Puddephat et al. (2001).
Whole T0 plants were regenerated from transgenic hairy root lines
containing T-DNA constructs with the B. oleracea ACC oxidase 1 and 2 and ACC
synthase cDNAs in both sense and antisense orientations. Five lines had been
transformed with sense constructs, 10 with antisense, and 3 with gus. It is difficult
to draw conclusions about whether these constructs had any effect on root
production without ethylene measurements. It is possible to argue that the
promoters were actively producing mRNA transcripts as shown by the GFP and
GUS assays, but impossible to determine whether this had any physiological
effect. There is an overwhelming body of literature suggesting that ethylene is
inhibitory to shoot formation, and that reducing it would enhance regeneration.
Inhibitors of ethylene production (AVG) and ethylene action (silver ions) have been
used to enhance shoot production of Brassica oleracea var. italica/gemmifera
(Sethi et al., 1990), Zea mays (Songstad et al., 1988), and Triticum aestivum
114
(Purnhauser et al., 1987) from callus. In Helianthus annuus, ethylene appeared to
enhance callus formation but inhibited shoot production (Robinson and Adams,
1987). However, the role of ethylene in cultured plant cells and tissues remains
unclear (reviewed by Kumar et al., 1998).
Pua and Chi (1993) reintroduced the ACC oxidase of B. juncea in antisense
orientation, producing lines that showed a greater capacity for shoot
morphogenesis than the controls. This was not the case in the present
experiments as actively growing tissue was the most important factor. Actively
growing tissue was found in lines with the gus control and ACC synthase
antisense constructs, which should be able to produce ethylene, but were more
amenable to regeneration. Pua and Chi (1993) associated the presence of
ethylene with culture browning, which is inhibitory to cell growth and/or
differentiation. Culture browning is a major problem for tissue culture, but was
necessarily related to ethylene in the current experiments, as this would have
selected for lines that down-regulated ethylene and not those of the control or
sense lines. This is not observed, so other factors may play a role in tissue
browning.
It was demonstrated that the ACC constructs had affected regeneration by
the time it took between inoculation of the explant and production of a mature
broccoli head. The lines transformed with the gus construct, which would not
manipulate ethylene, matured significantly faster (p<0.001), than the lines
containing the ACC constructs. It suggests that ethylene is required for normal
plant developmental growth.
Initially, explants were inoculated with A. rhizogenes strain LBA 9402
pRi1855::GFP (Spano et al., 1982, Puddephat et al., 2001) carrying the constructs
115
(chapter 2) on the binary plasmids. Transformants were selected for by ‘hairy root’
green fluorescence under long-wave UV light, and regenerated though tissue culture.
The rate of co-transformation of the host genome by the TL-DNA and binary T-DNA
was found to be 96% with the gus root lines. DNA was extracted from leaf material of
regenerants and checked for the presence/absence of the gfp gene by PCR. As the
gfp gene is located on the TL-DNA region it is used as a marker for identifying inserts
containing the rol genes. The PCR confirmed that the target DNA sequence was
present in the samples, although it may be argued that the DNA may have not
integrated into the plant genome or have been present as a result A. rhizogenes
contamination. The southern blots and hybridisation with specific DNA probes was
carried out to determine copy number and stable integration of the T-DNAs into the
plant host genome. The main limitation to this procedure was extracting DNA without
co-purifying charged polysaccharides. The Kirby method was most effective, and
although leaving traces of polysaccharides, was digested with the SstI restriction
endonuclease as shown in Figures 3.8 and 3.11. The impure DNA led to the diffusion
of some of the samples from the wells of the gels after loading, and also made it
difficult to equalise the amounts of DNA in each lane. However, data for gfp inserts
was obtained for 8 of the lines, showing that copy number ranged from 1-4 inserts.
The number of inserts did not appear to have any effect on the ‘hairy root’ phenotype
shown by the T0 regenerants (Table 3.7). For example, 28/00 aco1A 3 contained 4
gfp inserts and had a weak rol phenotype, whereas 5/00 aco2S 1 had a moderate
phenotype with only 1 insert. The effects of insert number on phenotype suggest
either position effects or possible silencing as the one copy of 5/00 aco2S 1 was
more active than the 3 copies in 28/00 aco1A 3. There is no data for the regenerants
with severe rol phenotype as the extraction did not produce very pure DNA.
116
The DNA extracted by the Kirby method was used to confirm presence of the
construct binary T-DNA in the plant genomes. A PCR was carried out with sequence
specific primers to the CaMV 35S promoter and cDNAs of the particular genes. All of
the primer pairs worked except for aco1S that produced a smear. The negative
results from 7/00 acsA 13 and 9/00 acsA 2 suggest no construct T-DNA insert, as
the DNA had been amplified with the gfp primers. Only 6 lines produced positive
bands for the constructs in the autoradiographs, and copy number ranged from 1-3.
The results from the autoradiographs did not always match the PCR data as the
Southern blots may have been improved with pure DNA at the same concentration.
However, this has shown that T-DNA insert copy number appears to range between
1 and 3 in the plant genome of GDDH33.
The A. rhizogenes-mediated transformation produced 18 T0 plants, all
derived from single cells. It is important that each cell of the transformants
contains the new construct for quantifying the impact of the transgenes on plant
physiology, including the analysis of ethylene and chlorophyll. Berthomieu et al.
(1994) produced a high frequency of chimeric plants with an A. tumefaciens
protocol that would have been undesirable. The impact of the rol genes was
severe in 4 of the 18 lines showing extremely wrinkled leaves, shortened
internodes and smaller heads with reduced chlorophyll. These characteristics,
termed the ‘hairy root syndrome’ by Tepfer (1990), have been attributed to certain
loci following knockout experiments. Wrinkled leaves and reduced internodal
distance are caused by rolA (Sinkar et al., 1988); rolB produces flower heterostyly
and abundant adventitious rooting (Cardarelli et al., 1987); while rolC results in
reduced apical dominance, altered leaf morphology and reduced seed production.
It is not uncommon to produce a variable phenotype when regenerating from A.
117
rhizogenes transformed plants. Christey et al. (1999) produced a range of
morphological differences with 20% normal phenotype, 40% moderate or slight
hairy root phenotype and 40% with severe phenotype.
The negative effects of rol genes on plant mortality have not previously
been reported. In these experiments, only 31% of the plants with severe hairy root
phenotype grew to maturity. Eighty-nine percent of plants with a normal phenotpye
and 67% with a moderate phenotype survived. The plants with severe phenotype
were also more susceptible to biotic and abiotic stresses in the glasshouse.
In total, there were approximately 6000 seeds recovered from 11 of the T0
lines. This is a very high yield considering the heads were removed for ethylene
and chlorophyll analysis (chapter 4) and floral organs had to grow from offshoots.
It is not surprising that there were no seeds recovered from plants with severe rol
phenotype as they had very small heads so even the offshoots were removed for
the analysis. It is difficult to compare seed production from different lines as there
were disparate plant numbers and varying degrees of decapitation. It does show
that the constructs and TL-DNA were not too inhibitory on seed production. Sixty-
seven progeny from 69 seeds of 28/00 aco1A 3 were tested for the presence or
absence of the gfp and aco1A DNA sequences by PCR. The autoradiographs had
shown that there were 4 TL-DNA and 2 construct T-DNA inserts. The largest group
of segregants were the double-positives with 56% suggesting a possible linkage.
The smallest group with 3% were the segregants that only contained an aco1A
insert. These data suggest that the TL-DNA and the construct T-DNA that have co-
transformed together are physically linked. It also shows, that given enough seeds
it is possible to segregate out the unwanted marker genes to leave phenotypically
118
normal transgenic plants that contain desirable characteristics through this
transformation protocol.
As the main aim of this part of the project was to transform GDDH33 and
regenerate transgenic plants, there were aspects that could have been improved.
If there had been more time available it would have been useful to improve the
transformation efficiency by enhancing the virulence of the bacteria, or T-DNA
integration. Researchers have used acetosyringone and mannopine along with
feeder layers to improve virulence (Henzi et al., 2000), but the main limiting factor
appears to be T-DNA integration (Gelvin, 2000). If there had been more transgenic
root available it would have been possible to take ethylene measurements, and
determine the impact on growth and regeneration. It would have also been
possible to add either substrates (ACC) or inhibitors of ethylene (silver ions/AVG)
to the media to obtain the optimum combination for both root growth and
regeneration. In future experiments, it might be worth using a broccoli post-harvest
head-specific promoter that does not interfere with root physiology such as the
senescence-related promoter used by Gan and Amasino (1995).
In these experiments, the A. rhizogenes co-integrate strain LBA 9402
pRi1855::GFP containing B. oleracea ACC oxidase 1 and 2 and ACC synthase in
sense and antisense orientations transformed the doubled haploid calabrese line
GDDH33 with an overall efficiency of 3.6 0.6%. Transformation efficiency was
greater with constructs that did not manipulate ethylene, and improved by
inoculating the cut surface soon after excision. The GFP reporter gene was an
effective marker for transgene expression in hairy roots. It was used with the gus
gene to produce 96% co-transformed roots. Reducing ethylene may have had an
adverse effect on root growth but beneficial on shoot regeneration, although the
119
most important factor for regeneration was that the tissue was actively growing.
Confirmation of transformation and stable integration of T-DNA inserts was
achieved by PCR and Southern blots. The copy numbers of these two species of
T-DNA ranged from 1-4 inserts, which did not relate to the extent of the severity of
the rol phenotpye. The rol genes had negative effects on 4 out of 18 T0 plants
reducing the size of plant and increasing mortality. It was possible to obtain viable
seed from 11 of the T0 plants and test some of them for segregation of T-DNAs. It
showed that the T-DNAs were probably physically linked, but it was possible to
produce marker-free, phenotypically normal plants with a desirable characteristic.
Given more time and root tissue, the efficiency of the processes may have
been improved, but the main aim of producing transgenic plants containing the
ACC cDNAs in sense and antisense orientations to be measured for post-harvest
ethylene production was achieved.
120
4.0 Analysis of broccoli bud
post-harvest ethylene production
and chlorophyll levels
121
4.1 Introduction
Broccoli is one of the most successful ‘new’ crops in the United
Kingdom. The crop value of home produced marketed broccoli has doubled
from £18.9m to £44.9m in the last 10 years (DEFRA statistics, 2000), where
there has been an overall decline of marketed Brassicas. Broccoli crops of the
same variety treated in the same way after harvest vary in their post-harvest
performance (D. Pink, pers. comm. HRI-Wellesbourne). This lack of
predictability results in the major retailers specifying a product life of only two
days from delivery to ‘purchase before date’, and a further two days ‘best
consumed by’ in an attempt to ensure a ‘quality’ product for the consumer.
This short post-harvest life results in significant waste for the retailer (a cost
borne by the consumer) and consumers themselves.
Broccoli is a floral vegetable, harvested when the flowering heads are
immature and growing rapidly (King and Morris, 1994a,b). Each broccoli floret
is composed of male and female reproductive structures surrounded by
immature petals and enclosed within chlorophyll containing sepals (Pogson et
al., 1995a). Post-harvest senescence is characterised by floret chlorophyll
loss resulting in the yellowing of sepals (Wang, 1977). Sepal yellowing
commences between 24 and 48h after harvest (20oC) and is essentially complete by 96h (Tian et al., 1994). It is a relatively late event in the post-
harvest senescence of broccoli preceded by major losses of sugars, organic
acids and proteins (King and Morris, 1994).
Endogenous ethylene production is suggested to have an important
role in the colour change of stored broccoli (Hyodo et al., 1994; King and
122
Morris, 1994). Exogenous ethylene treatment has been shown to accelerate
yellowing of broccoli (Aharoni et al., 1985) and the effect has been attributed
to stimulation of endogenous ethylene production (Makhlouf et al., 1992) and
increased tissue sensitivity to ethylene (Tian et al., 1994). Furthermore,
ethylene biosynthetic inhibitors such as aminoethoxyvinylglycine and
inhibitors of ethylene action (silver ions and methylcyclopropene) delay
chlorophyll loss (Wang, 1977; Ku and Wills, 1999).
In plants, ethylene production is induced during several developmental
stages, including fruit ripening, seed germination, leaf and flower senescence,
and abscission. It is also induced by external factors, such as wounding,
anaerobiosis, viral infection, auxin treatment, chilling injury and drought
(Abeles, 1973; Yang and Hoffman, 1984). Ethylene biosynthesis occurs via
the enzymes ACC synthase and ACC oxidase, which catalyse the
conversions of S-adenosylmethionine to 1-aminocyclopropane-1-carboxylic
acid (ACC) and ACC to ethylene, respectively (see Figure 1.2) (Yang and
Hoffman, 1984). In this pathway, the conversion of SAM to ACC is generally
regarded as the rate-limiting step, although there is a large body of evidence
indicating that regulation of ACC oxidase is also important (Yamamoto et al.,
1995; Barry et al., 1996).
Detailed studies of the physiological changes occurring within
individual broccoli florets after harvest suggest that the reproductive structures
of the floral buds may exert some influence over chlorophyll loss. An increase
in ACC oxidase activity and ethylene production occurs in the reproductive
structures and removal of these significantly reduces the rate of sepal
yellowing (Tian et al., 1994). Senescence in carnations and orchid flowers is
123
associated with increased ethylene production arising from concomitant
increases in both ACC synthase and ACC oxidase (Woodson et al., 1992;
O’Neil et al., 1993). Tissues producing ethylene also become more sensitive
to ethylene due to increased expression of ethylene receptors (Payton et al.,
1996).
Pogson et al. (1995a) isolated two cDNA clones, ACC oxidase 1 (aco1)
and ACC oxidase 2 (aco2) from a Brassica oleracea L. var. italica cDNA
library, derived from florets 48h post-harvest. The lengths of the aco1 and
aco2 were 1237 and 1232bp, respectively. They shared nucleotide identity of
83% in their putative translated regions (86% amino acid identity) and 48% in
the 3’ untranslated regions. Transcripts were found at low levels in whole
florets at the time of harvest and increased markedly in abundance after
harvest. aco1 transcript abundance also increased in sepals after harvest and
in excised yellowing leaves. Transcripts corresponding to aco2 were found
exclusively within the reproductive structures and accumulated to high
abundance post-harvest. Pogson et al. (1995b) also isolated an ACC
synthase (acs) cDNA clone from B. oleracea with an apple cDNA probe. The
transcript abundance of ACC synthase did not change in florets 0, 24, 48, 72h
after harvest and was detected in similar abundance in green and senescing
leaves. ACC synthases are known to belong to multigene families and there
at least three different classes of ACC synthase in tomato (Rottman et al.,
1991). ACC synthase sequences are very divergent at the nucleotide level
(Yang and Oetiker, 1998), and it is possible that the cDNA apple probe used
by Pogson et al. (1995b) may not have hybridised to all of the cDNAs
encoding for ACC synthase enzymes. At least two other ACC synthase
124
cDNAs have been identified (Gonzalez and Botella, 2001) in post-harvest
broccoli, which share 38% and 42% nucleotide identity with the clone isolated
by Pogson et al. (1995b).
Gene silencing with sense genes has been an important method for
down-regulating the expression of endogenous plant genes (for review see
Meyer and Saedler, 1996). However, the efficiency of silencing is
unpredictable and the underlying mechanisms are unknown (Hamilton et al.,
1998). It is possible that sense constructs may up-regulate genes causing an
over-expression of transcripts that can also reveal information about gene
function. Hamilton et al. (1990) reduced ethylene synthesis in tomato fruit by
93%, by transformation with pTOM13 an ACC oxidase in antisense orientation
in relation to a cauliflower mosaic virus (CaMV) 35S promoter. Henzi et al.
(1999a, b) transformed Brassica oleracea L. var. italica with the same
pTOM13 construct. There was generally greater production of ethylene in the
flowers at 26h post-harvest in the transgenic lines than in the untransformed
control and therefore, the construct did not reduce post-harvest ethylene
production or chlorophyll loss. The tomato ACC oxidase only shared 70%
nucleotide identity to the native B. oleracea ACC oxidases, which may not
have been enough for effective gene-silencing.
The B. oleracea L. var. italica ACC oxidase 1 and 2 and ACC synthase
cDNAs have been obtained from Pogson et al. (1995), and ligated between a
CaMV 35S promoter and nopaline synthase (nos) terminator in sense and
antisense orientations (chapter 2). Additionally, a -glucuronidase gene (gus)
(Jefferson et al., 1987) was ligated between the same promoter and
terminator for a control. Constructs were inserted into the T-DNA region of
125
pSCV1.0 (Biogemma) and introduced into the Agrobacterium rhizogenes co-
integrate strain LBA 9402 pRi1855::GFP by electroporation (chapter 2).
Doubled haploid genotypes possess uniform genetic backgrounds,
which are useful for breeding, gene mapping and quantifying an additional
trait by plant transformation. In a field trial conducted at HRI, the post-harvest
shelf-life of broccoli cultivars was assessed. The mean time for bud yellowing
of GDDH33 was 46h 19min, 53h 23min for a doubled haploid line of Shogun
and 81h 28min for Marathon F1 (D. Pink, pers. comm. HRI-Wellesbourne). As
GDDH33 had the poorest shelf-life performance it was selected as a model to
test the relation between post-harvest ethylene production and chlorophyll
loss.
Explants of GDDH33 were inoculated with Agrobacterium rhizogenes
co-integrate strain LBA 9402 pRi1855::GFP and transformants were selected
to be regenerated into whole plants (see chapter 3). Seventeen lines were
regenerated: 3 aco1 antisense, 2 aco1 sense, 2 aco2 antisense, 2 aco2
sense, 4 acs antisense and 1 acs sense and 3 gus.
The aim of the research was to a) determine the non/post-harvest
physiological changes of ethylene and chlorophyll in the untransformed
GDDH33 broccoli buds; b) test the post-harvest broccoli buds of T0 plants for
ethylene production and chlorophyll loss; c) to determine if is there a
relationship between post-harvest ethylene production and chlorophyll loss in
senescing buds; and d) to assess whether a transgenic method could produce
broccoli florets that retain chlorophyll for a greater time and increase product
shelf-life.
126
It was necessary to develop assays to measure ethylene and
chlorophyll from broccoli florets. Previous studies have measured ethylene
and chlorophyll from whole heads, florets and branchlets (King and Morris,
1994a,b; Pogson et al., 1995a). Tian et al. (1994) used a more elegant
approach by focusing on the floral buds. These were the organs responsible
for post-harvest production of ethylene and experienced chlorophyll loss in the
sepals.
127
4.2 Materials and Methods 4.2.1 Harvested broccoli heads
GDDH33 is a doubled-haploid line derived from the calabrese cultivar
Green Duke through anther culture. It was grown from seed to head maturity
in the glasshouses of HRI Wellesbourne (15oC min with forced air at 18oC,
day length extension to 14h by high pressure sodium lamps) throughout the
year in 15cm pots with non-sterile compost (a 80:20 mix of Levington M2 and
vermiculite). Transgenic lines were produced by transformation of GDDH33
with A. rhizogenes co-integrate strain LBA 9402 pRi1855::GFP carrying
Brassica oleracea ACC oxidase 1 and 2 and ACC synthase cDNAs (Pogson
et al., 1995a,b) in sense and antisense orientations in relation to a CaMV 35S
promoter and nos terminator and a gus gene construct (Jefferson et al., 1987)
on the binary vector T-DNA. Seventeen transgenic lines were regenerated for
analysis (shown in Table 4.1). They had been regenerated from hairy roots,
though a callus phase into plantlets. These were grown to maturity in air
conditioned container glasshouse compartments set to maintain 22oC by
forced air ventilation with a minimum 14 hours photoperiod at HRI
Wellesbourne, throughout the year.
128
Table 4.1 Transgenic lines regenerated by tissue culture
Experiment
No. ACC cDNA* Sense/
Antisense
Explant no.
producing gfp root
28/00 aco1 A 3
34/00 aco1 A 1
34/00 aco1 A 4
32/00 aco1 S 3
33/00 aco1 S 4
3/00 aco2 A 2
33/00 aco2 A 3
5/00 aco2 S 1
26/00 aco2 S 1
7/00 acs A 12
7/00 acs A 13
9/00 acs A 1
9/00 acs A 2
7/00 acs S 7
5/00 gus - 1
33/00 gus - 3
34/00 gus - 2
*aco1 =ACC oxidase 1, aco2= ACC oxidase 2, acs = ACC synthase, gus= glucuronidase
At maturity, whole heads were harvested in the morning at 9-10am,
and transported to the laboratory within 20min. Heads were surface sterilised
with 100ppm NaOCl as described by Rushing (1990) and placed into a 20oC
incubator in the dark. At 0, 24, 48, 72 and 96h post-harvest, samples of 0.2g
floral buds were removed from the broccoli heads with a scalpel. There were
12 different head replicates for each time sample, for the untransformed
129
GDDH33 and 3 replicates for the T0 lines. At head maturity, bud samples
were also taken directly from the untransformed control plants growing in situ.
These samples were termed ‘non-harvested’ as the heads from which the
buds were removed were still attached to the stem. Floral buds were removed
from 12 intact broccoli heads with a scalpel at 0, 24, 48, 72 and 96h.
4.2.2 Ethylene Measurements
A serial dilution of ethylene was prepared to calibrate gas
chromatograph readings with ethylene concentrations. A 25ml flask was
submerged in water to displace its air. It was filled with ethylene (BOC) to
displace the water and fitted with a gas-tight subaseal (Fisher). Three 1L
volumetric flasks were filled with water, which was concurrently displaced with
nitrogen (BOC) and the flasks sealed. A volume of nitrogen was removed and
replaced with saturated ammonium sulphate (Sigma) to leave a final volume
of 1L of nitrogen. A 1ml volume of ethylene was taken from the 25ml flask and
added to the first volumetric flask with 1L nitrogen to make a 1:1000 dilution.
This was mixed and left for 30 minutes. A 1ml sample was taken from this
flask and added to the next to make a 1:1000 000 dilution. Twelve samples
were taken from the 1ppm flask to calculate an average, for standardising
ethylene measurements from the broccoli buds.
At each sample point, 0.2g (fresh weight) of floral buds from the
broccoli heads were placed into a 25ml glass specimen bottle and fitted with a
subaseal (see Figure 4.1b). Sealed bottle, with buds, were placed in a 20oC
incubator in the dark for 1h. Three 1ml gas samples were then taken with a
gas-tight fixed-needle glass syringe and injected into a gas chromatograph
130
(GC, Analytical Instruments Model 93). The GC was fitted with a flame
ionisation detector containing a 2.0mm diameter stainless steel column
packed with Poropak Q (80-100 mesh) and a Spectra Physics integrator (see
Figures 4.1a,c). The oven temperature was kept constant at 100oC while the
column was kept at 60oC. In these conditions it took ethylene 1.3min to pass through the column. A value was recorded which could then be converted to
an amount of ethylene using the calibration.
4.2.3 Chlorophyll Measurements
A system to measure chlorophyll levels was devised. An experiment
was set up to test the speed of chlorophyll diffusion from broccoli buds into
methanol. Three replicates of 0.2g broccoli buds were added to 5ml of
methanol in 10ml round-bottomed tubes in the dark at 20oC (see Figure 4.4).
Samples of 100 l were taken after 60min for each, up to 480min. A second
experiment was set up to observe the effects of light on chlorophyll. Two
broccoli bud samples of 0.2g per head were taken from 6 heads and placed in
5ml methanol, in 10ml round-bottomed tubes. One treatment was placed in
the dark at 22oC, while the other was kept in the light (mix of 70 W white and
65/80 W gro-lux fluorescent tubes providing an irradiance of 80 mol m-² s-¹) at
22oC. Chlorophyll amounts were measured for all samples according to Hipkins and Baker (1986).
All chlorophyll measurements for untransformed and transgenic lines of
GDDH33 were standardised. The 0.2g of floral broccoli buds from the
ethylene measurements were transferred to a 10ml round-bottomed tube.
Five millilitres of methanol (BDH analaR grade) were added to the buds and
131
left overnight at 20oC in the dark. A 1ml sample was added to a cuvette and
measured using a single beam spectrophotometer, 10mm cell with
absorbance at 645 and 663nm. The values were converted to microgram
chlorophyll per gram fresh weight of broccoli buds (see Table 4.2).
Table 4.2 The concentration of chlorophyll (Chl) in the cuvette was calculated using simultaneous equations, based on the absorbance readings at 645 and 663nm (Hipkins and Baker, 1986).
Chl a 12.7 A663 - 2.69 A645
Chl b 22.9 A645 - 4.68 A663
Chl (total) 20.2 A645 + 8.02 A663
4.2.4 Measuring dry weight of broccoli buds
A 9cm diameter Whatman 1 filter paper was folded to fit into a 10cm
diameter pyrex funnel. The 5ml of methanol and broccoli buds were decanted
onto the filter paper. The tube was then filled with 10ml of water, which was
also decanted onto the filter paper. Only the buds remained on the filter paper
after this procedure. The filter paper was then placed in a 5cm by 5cm brown
paper envelope and dried at 60oC for 2 days. Dried buds were weighed on a
microbalance and values recorded. These values were used to calculate
ethylene production nanolitre per gram dry weight of broccoli buds per hour.
4.2.5 Data Analysis
Statistical analyses were carried out on the data. The comparison of
sample means was performed with Students T-Test, and calculating the
probability of regression with an Anaysis of Variance (ANOVA) test.
132
(a) (b)
(d)
(c)
Figure 4.1 (a) Gas chromatograph; (b) Glass container with broccoli buds evolving ethylene; (c) Injecting 1ml air sample taken from (b); (d) Chlorophyll diffusing from broccoli buds in 5mls methanol.
133
4.3 Results 4.3.1 Tissue to be assayed
The first experiments were designed to provide quantitative assays for
measuring ethylene and chlorophyll from broccoli heads. Whole heads were
the first tissue samples to be used but these gave inconsistent ethylene
values, because the stalks were of different weights. The whole heads also
gave off volatiles that were interfering with the column of the gas
chromatograph, so that ethylene measurements were inconsistent between
samples. It was decided to focus on broccoli buds for the ethylene and
chlorophyll assays as these provided consistent ethylene results and could
easily be used for chlorophyll assays.
4.3.2 Controls for standardising the analysis of ethylene and chlorophyll 4.3.2.1 Calculating ethylene concentrations
The GC value for 1ppm ethylene was 7419 93. The amount of
ethylene produced (nl) for 1g (dry weight) of broccoli buds per hour is
calculated from the expression: [mean ethylene value from 3 measurements/
ethylene value for 1ppm (7419)] x weight conversion factor (converts dry bud
weight to 1g i.e 1/value) x volume of air in specimen tube (25mls).
At 20oC, the rate of diffusion of chlorophyll from broccoli buds in methanol was time dependent (see Figure 4.2). In this experiment, after 300
minutes the rate of diffusion began to level off as all of the chlorophyll had
diffused into the methanol. The total quantity of chlorophyll per sample is an
absolute value, and represents the amount in the broccoli buds at any given
134
Chl
orop
hyll
(ug
in s
olut
ion/
g bu
d)
time. As this could only be obtained by leaving the buds for at least 480min, it was decided to leave the broccoli buds overnight, in the dark.
800
700
600
500
400
300
200
100
0
0 60 120 180 240 300 360 420 480
Time (min)
Figure 4.2 The rate of diffusion of chlorophyll at 20oC in the dark, from 0.2g broccoli buds in 5ml methanol. Sample 1( ) , sample 2 ( ), sample 3 ( ).
4.3.2.2 Analysis of pre/post harvest ethylene production and chlorophyll levels
in broccoli buds of non-transformed GDDH33
In Figure 4.3, the basal amount of ethylene produced by non-harvested
GDDH33 broccoli buds was 44 1nl/g bud/h. After harvesting, an increase in
ethylene production was observed. There was little physiological change at 0h
between harvested (61 1nl/g bud/h) and non-harvested buds (44 1nl/g
bud/h). The major difference occurred at 24h, when the harvested buds
produced 176 18nl/g bud/h ethylene, significantly more than the non-
harvested buds with 44 1nl/g bud/h [t(df)= 7(24) p<0.005]. This peak is
clearly a response to harvest, and stays high for 48h (116 13nl/g bud/h)
before declining to the basal amount at 72h and 96h.
135
Eth
ylen
e (n
l/g b
ud/h
)
250
200
150
100
50
0 0 24 48 72 96
Time (h)
Figure 4.3 Ethylene production by harvested/non-harvested untransformed GDDH33 broccoli buds. Harvested buds ( ), and non-harvested ( ), n=1x12.
In Figure 4.4, the non-harvested broccoli buds had no detectable
change in chlorophyll levels over this period of time with an overall mean of
651 50 g/g bud. The loss of chlorophyll in the harvested buds is a direct
physiological response to the removal of the head. As with Figure 4.3, there
was insufficient time for there to be much of a difference between non-
harvested and harvested buds at 0h. The major reductions of chlorophyll
levels occurred between 0h and 72h. Chlorophyll loss occurred very rapidly,
initially at 24h, 26% was lost compared to 0h, and by 96h over 80%
chlorophyll was lost.
136
Chl
orop
hyll
(ug/
g bu
d)
900
800
700
600
500
400
300
200
100
0
0 24 48 72 96
Time (h)
Figure 4.4 Chlorophyll levels of harvested/non-harvested untransformed GDDH33 broccoli buds. Harvested buds ( ), non-harvested ( ), n=1x12.
4.3.3 Analysis of post-harvest ethylene production and chlorophyll levels of
non-transformed GDDH33 and the transgenic lines
Tables 4.3 and 4.4 contain the underlying data for the Figures 4.5-4.14
as reference in the rest of this section. The data for ‘construct lines’ has been
grouped together (for each particular construct) as all of the lines have
performed similarly, except for ACC synthase antisense. The ‘total’ ethylene
production is defined better as an ‘ethylene production index’ as it is not the
total amount of ethylene that the broccoli buds have produced post-harvest,
but the sum of the recorded values over 96h. The percentage of chlorophyll
loss in Table 4.4 is the difference between the value at 96h and that at 0h,
expressed as a percentage.
137
Table 4.3 Ethylene levels of GDDH33 buds post-harvest as nanolitres per gram bud (dry weight) per hour. The means and SEs for each time point were calculated with numbers of lines (n) and numbers of replicates. For the constructs these were: untransformed GDDH33 n=1x12; and transformed with gus, n=3x3; aco1 antisense, n=3x3; aco1 sense, n=2x3; aco2 antisense, n=2x3; aco2 sense, n=2x3; acs antisense, n=4x3; and acs sense, n=1x3.
Means S.E.s Constructs Time (h) Time (h)
0 24 48 72 96 Total 0 24 48 72 96
Un-t 61.9 176.1 116.0 52.9 37.7 444.7 10.5 18.1 22.4 7.7 16.4
gus 56.9 114.6 99.7 42.9 52.3 366.4 9.6 26.1 25.4 19.2 11.0
aco1A 35.8 31.9 59.2 46.7 48.3 221.9 6.8 9.0 15.0 9.8 12.4
aco1S 23.4 59.3 49.8 30.3 31.0 193.7 3.7 12.3 12.0 10.4 12.0
aco2A 10.6 44.9 32.2 47.4 9.8 144.8 5.0 19.3 6.5 12.0 4.8
aco2S 52.6 49.1 39.9 54.0 65.7 261.3 8.0 8.1 5.4 11.5 13.5
acsA 59.0 79.9 74.8 92.1 48.9 354.8 10.9 20.1 11.3 15.4 8.6
acsS 245.6 7.3 79.4 77.1 30.9 440.3 65.2 4.7 39.7 16.6 30.9
Table 4.4 Chlorophyll values of GDDH33 buds post-harvest as micrograms per gram bud (fresh weight). The means and SEs for each time point were calculated with numbers of lines (n) and numbers of replicates. For the constructs these were: untransformed GDDH33 n=1x12; and transformed with gus, n=3x3; aco1 antisense, n=3x3; aco1 sense, n=2x3; aco2 antisense, n=2x3; aco2 sense, n=2x3; acs antisense, n=4x3; and acs sense, n=1x3.
Means S.E.s
Constructs Time (h) Time (h)
0 24 48 72 96 %loss 0 24 48 72 96
Un-t 721.5 535.8 229.3 145.7 140.9 80.5 21.4 36.5 19.3 18.9 14.4
gus 409.7 387.9 338.7 164.4 159.1 61.2 48.7 44.1 38.7 48.9 34.2
aco1A 292.3 307.4 290. 8 212.7 175.8 39.9 18.6 17.9 21.6 32.5 32.6
aco1S 355.8 379.8 379.2 305.1 361.7 -1.7 19.9 27.3 42.5 54.8 86.9
aco2A 365.8 399.1 273.5 293.3 299.7 18.1 52.2 35.6 26.4 20.6 45.8
aco2S 279.3 277.4 210.6 202.6 146.8 47.4 32.5 27.8 27.9 43.1 37.6
acsA 268.1 276.5 178.8 142.4 91.2 66.0 38.42 40.9 18.9 31.5 21.5
acsS 200.4 247.1 294.8 164.5 161.6 19.3 28.6 20.4 109.2 60.1 121.1
138
Figure 4.5 includes the post-harvest data of Figures 4.3 and 4.4 in one
graph. It demonstrates the post-harvest elevation of ethylene production with
rapid chlorophyll loss in untransformed GDDH33.
The gus control plants showed the impact of transformation on
ethylene and chlorophyll levels in GDDH33 broccoli buds. The transformed
plants in Figure 4.6 have reduced ethylene and chlorophyll levels in
comparison to the untransformed control. The ethylene peak at 24h in Figure
4.5 of 176nl/g bud/h is reduced to 116nl/g bud/h in the transformed plants, but
this is not significant [t(df)= 2(22), p>0.005], and the other values are
comparable. The transformed plants produced 18% less ‘total’ ethylene over
the time period compared to the untransformed control, as a result of the
reduced peak. At 0h, there is a 43% reduction of chlorophyll from 721 g/g bud
in the untransformed to 409 g/g bud in the transformed plants. However, the
rate of chlorophyll loss in the transformed plant is also reduced and occurs
mostly between 48 and 72h, after the ethylene peak. After 48h, there has only
been 18% chlorophyll loss in the transformed plant, whereas there has been
68% in the untransformed. It is clear that the effects of transformation and
regeneration have reduced ethylene production and the starting amount of
chlorophyll, and also reduced chlorophyll loss.
In Figure 4.6, the effect of the transformation and regeneration reduced
total ethylene production of GDDH33 buds by 18%. In Figure 4.7, the plants
transformed with aco1 antisense construct produced 50% less ‘total’ ethylene
than the untransformed control and 32% less than the gus controls. At 24h,
where the untransformed buds produced 176nl/h ethylene, the plants
transformed with aco1 antisense produced only 32nl/h, an 82% reduction and
139
significantly less [t(df)= 6(17), p<0.005], than the gus control. aco1 antisense
constructs reduced ethylene to the basal levels required for normal physiology
shown in Figure 4.3. Chlorophyll loss started to occur at 48h, but over the
whole time period lost only 40%, significantly less than the untransformed
control [t(df)= 4(20), p<0.005], but not significantly less than 62% in the gus
control. It is clear both the untransformed and gus controls started with more
chlorophyll, and therefore had more to lose. However, the aco1 antisense
plants had more chlorophyll at 96h than both, so must have retained a much
greater proportion.
Transformation of GDDH33 with the aco1 sense constructs reduced
ethylene and prevented chlorophyll loss (Figure 4.8), over this time period.
Ethylene evolved from the broccoli buds produced the same shape curve as
the controls with a peak at 24h. The aco1 sense lines produced significantly
less ethylene at 24h than the untransformed control [t(df)= 4(19), p<0.005].
This peak was 49% of the gus control, but was not significantly different [t(df)=
1.67(13), p>0.005].
The quantity of chlorophyll in aco1 sense lines (355 20 g/g bud) at 0h is similar to the gus control (409 48 g/g bud). However, by 96h
chlorophyll loss of buds in the gus line is significantly greater [t(df)= 2.55(19),
p<0.005] than in the aco1 sense line where there is no chlorophyll loss.
The curves in Figure 4.9 are similar to those of Figures 4.6 and 4.8.
Ethylene has been significantly reduced by aco2 antisense constructs in
comparison to the untransformed control significant [t(df)= 4(18), p<0.005] to
non-harvest basal levels. Between 0h and 96h, chlorophyll loss has been
significantly reduced from 61% in the gus controls to only 18% with the aco2
140
antisense lines [t(df)= 2.99(13), p<0.005]. Again it shows that by manipulating
ACO enzymes through mRNA transcript suppression, post-harvest ethylene
production and chlorophyll loss have been reduced.
In Figure 4.10, the aco2 sense constructs also reduced post-harvest
production of ethylene to basal levels. At 24h there was significantly less
ethylene production than the untransformed control [t(df)= 2.29(17), p<0.005].
There was greater chlorophyll loss (48%) from these buds than the other ACO
lines at 96h but, this was still significantly less than the untransformed control
[t(df)= 2.89(7), p<0.005] and less than the gus control plants (61%).
In Figure 4.11, the ACC synthase antisense constructs did not down-
regulate post-harvest ethylene production as much as the ACO constructs.
Total ethylene production was much higher (354nl/g bud/h), similar to the gus
control plants (366nl/g bud/h). However, the ethylene peak from the gus
plants occurred between 24-48h, producing 114 and 99nl/g bud/h ethylene,
respectively. In the ACC synthase antisense lines the values at these time
points are 79 and 74nl/g bud/h, respectively, and the peak has been shifted to
72h. It is difficult to observe this with Figure 4.11, so Figure 4.12 has been
included. There was 66% chlorophyll loss over the time period, which is
slightly higher than the gus control (61%) and not significantly less than the
untransformed control [t(df)= 1.33(12), p<0.005].
All nine lines transformed with the aco1 and aco2 sense and antisense
constructs produced relatively similar ethylene curves post-harvest. There
was more variation between lines transformed with acs antisense. The mean
values in Figure 4.11 do not take into account such variation, so Figure 4.12
has been provided to illustrate this disparity. In Figure 4.12, 9/00 acsA 1 has
141
produced a similar post-harvest ethylene curve to the untransformed GDDH33
and gus controls. Either acs is not important in post-harvest ethylene, or the
construct has not down-regulated the native gene. The latter seems more
plausible, as the peak of ethylene production has shifted from 24h (126nl/g
bud/h) in 9/00 acsA 1 to 72h (143nl/g bud/ h) in 7/00 acsA 12. If 9/00 acsA 1
has not been manipulated by the acs antisense construct, then 7/00 acsA 12
must have been to produce such a different curve. It follows that the peak at
24h, present in the controls and 9/00 acsA 1 of Figure 4.12, must be
dependent on acs and production of ACC, as this is absent in 7/00 acsA 12.
The line transformed with the ACC synthase sense construct in Figure 4.13 produced the most distinctive post-harvest ethylene curve out of all of the
lines. Ethylene production at harvest (0h) is very high (245nl/g bud/h). In
Figures 4.3 and 4.4 there was very little physiological change between the
non-harvested/harvested buds at 0h. As there is a very high amount of
ethylene production at 0h, it suggests that ethylene is being produced at very
high levels in growing broccoli buds of the ACC synthase sense line.
However, between 0 and 24h there is a very rapid post-harvest decline in
ethylene production. These are the highest and lowest ethylene values of all
18 lines, and occur over a very short time. The peak recorded at 24h for the
untransformed and gus controls has been reduced by 96 and 94%,
respectively, suggesting that ACC synthase plays a role in the post-harvest
ethylene peak. Similar to ACC synthase antisense, there is recovery between
24 and 96h.
There has been a large decrease in ethylene production at 24h and
post-harvest chlorophyll levels actually increased between 0-48h from 200 to
142
294 g/g bud (see Table 4.4). The overall post-harvest chlorophyll loss was
not significantly less than the gus control [t(df)= 1.44(12), p<0.005] as this
occurred parallel to the recovery of ethylene production 48-72h where the
broccoli buds produced 79.38 and 77.09 nl/g bud/h ethylene, respectively.
In Figure 4.14, there is a positive correlation between the ‘total’
ethylene production and chlorophyll loss of post-harvest broccoli buds. As
ethylene production increases, chlorophyll loss increases. The significance of
the regression was tested with the F-distribution by comparing the ratio of the
regression and the residual mean squares. The percentage chlorophyll loss
response variate had a significant dependence [F(df)= 0.0015(7)] on the total
ethylene production variate. Total amount of ethylene production appears to
be the most important factor for chlorophyll loss. The untransformed GDDH33
is characterised by a large peak (176nl/ g bud/h) at 24h. Its rate of chlorophyll
loss was not significantly different to the ACC synthase antisense lines, which
peaked at 72h. If the timing of the ethylene peak was important, this would not
be the case. The transgenic lines that have reduced ethylene have also
reduced chlorophyll loss throughout the time period. It is clear in Figure 4.14
that reducing total ethylene production to less than 200nl/g bud/h is beneficial,
to decrease chlorophyll loss to only 20%. The ‘total’ amount of basal ethylene
production from non-harvested buds was 200nl/g bud/h from 0-96h.
Therefore, if post-harvest ethylene production is not greater than the basal
level, chlorophyll loss is significantly reduced. There was 80% chlorophyll loss
in the untransformed control with over double the amount of ethylene
production by 96h at 20oC.
143
In Figure 4.15, the post-harvest senescence of GDDH33 broccoli
heads from 0h-96h is shown visually. At 72h, bud yellowing can be observed
in the untransformed control, where there has already been 68% chlorophyll
loss. Chlorophyll loss through bud yellowing can be seen more clearly at 96h
where there has been 80% chlorophyll loss. There is no visible yellowing of
the head transformed with 3/00 aco2A 2 antisense constructs. However,
GDDH33 produces very small heads that are more susceptible to water loss,
and flaccidity cannot be prevented. It appears in the 3/00 aco2A 2 line that the
buds at 96h are larger than at 0h. If this is the case, the buds in the 3/00
aco2A 2 must have kept growing post-harvest.
144
Chl
orop
hyll
(ug/
g bu
d)
Chl
orop
hy ll
(ug/
g bu
d)
Ethy
lene
(nl/g
bud
/h)
Ethy
lene
(nl/g
bud
/h)
800
700
600
500
400
300
200
100
0
0 24 48 72 96
Time (h)
250 200 150
100 50 0
Figure 4.5 Post-harvest ethylene production and chlorophyll levels of buds from untransformed GDDH33. Chlorophyll levels ( ), ethylene production ( ), n=1x12.
800
700
600
500
400
300
200
100
0
0 24 48 72 96
Time (h)
250 200 150
100 50 0
Figure 4.6 Post-harvest ethylene production and chlorophyll levels of buds from GDDH33 transformed with gus control constructs. Chlorophyll levels ( ), ethylene production ( ), n=3x3.
145
Chl
orop
hyl
l (ug
/g b
ud)
Chl
orop
hyll
(ug/
g bu
d)
E th
ylen
e (n
l/g b
ud/
h)
Ehy
le ne
(nl/g
bud
/h)
800
700
600
500
400
300
200
100
0
0 24 48 72 96
Time (h)
250
200
150
100
50
0
Figure 4.7 Post-harvest ethylene production and chlorophyll levels of buds from GDDH33 transformed with ACC oxidase 1 antisense constructs. Chlorophyll levels ( ), ethylene production ( ), n=3x3.
800
700
600
500
400
300
200
100
0
0 24 48 72 96
Time (h)
250
200 150
100 50
0
Figure 4.8 Post-harvest ethylene production and chlorophyll levels of buds from GDDH33 transformed with ACC oxidase 1 sense constructs. Chlorophyll levels ( ), ethylene production ( ), n=2x3.
146
Chl
orop
hyll
(ug/
g bu
d)
Cho
roph
yll (
ug/g
bud
)
Eth
ylen
e (n
l/g b
ud/h
) Et
hyle
ne (
nl/g
bud
/h)
800
700
600
500
400
300
200
100
0
0 24 48 72 96
Time (h)
250
200
150 100
50
0
Figure 4.9 Post-harvest ethylene production and chlorophyll levels of buds from GDDH33 transformed with ACC oxidase 2 antisense constructs. Chlorophyll levels ( ), ethylene production ( ), n=2x3.
800
700
600
500
400
300
200
100
0
0 24 48 72 96
Time (h)
250 200 150
100 50 0
Figure 4.10 Post-harvest ethylene production and chlorophyll levels of buds from GDDH33 transformed with ACC oxidase 2 sense constructs. Chlorophyll levels ( ), ethylene production ( ), n=2x3.
147
Ethy
lene
(nl/g
bud
/h)
Chl
orop
hyll
(ug/
g bu
d)
Eth
ylen
e (n
l/g b
ud/h
)
800
700
600
500
400
300
200
100
0
0 24 48 72 96
Time (h)
250
200 150
100 50
0
Figure 4.11 Post-harvest ethylene production and chlorophyll levels of buds from GDDH33 transformed with ACC synthase antisense constructs. Chlorophyll levels ( ), ethylene production ( ), n=4x3.
250
200
150
100
50
0 0 24 48 72 96
Time (h) Figure 4.12 Post-harvest ethylene production of buds from two lines of GDDH33 transformed with ACC synthase antisense constructs. These are 9/00 acsA 1 ( ), and 7/00 acsA 12 ( ), n=3.
148
Tota
l chl
orop
hyll
loss
(%)
Chlo
roph
yll (
ug/g
bud
)
Ethy
lene
(nl/g
bud
/h)
1000
900
800
700
600
500
400
300
200
100
0
0 24 48 72 96
Time (h)
350 300
250 200 150 100
50 0
Figure 4.13 Post-harvest ethylene production and chlorophyll levels of buds from GDDH33 transformed with ACC synthase sense construct. Chlorophyll levels ( ), ethylene production ( ), n=1x3.
90 80 y = 0.2386x - 23.041
70 R2 = 0.8351
60
50
40
30
20
10
0 0 100 200 300 400 500
Total ethylene production (nl/g bud)
Figure 4.14 Relationship between post-harvest ethylene production and chlorophyll loss of buds from transformed and untransformed GDDH33. The markers ( ) represent all of the lines in Figures 4.5-4.13. The ACC sense line has been omitted as the chlorophyll levels increased for the first three ethylene values, and cannot be characterised by chlorophyll loss. Total ethylene production is the amount from each time point added together. The percentage chlorophyll loss is the difference between the starting chlorophyll value at 0h to that at 96h.
149
Figure 4.15 Broccoli heads post- harvest. Untransformed GDDH33 (left) and transformed 3/00 aco2A 2 antisense (right). From top to bottom: 0, 24, 48, 72 and 96h post- harvest broccoli heads.
(0) (24)
Untransformed 3/00 aco2A 2
(48)
(72)
(96)
150
4.4 Discussion
Broccoli buds were chosen for assaying ethylene and chlorophyll, as
they were the main organs responsible for post-harvest ethylene production
and experienced chlorophyll loss in the sepals (Tian et al., 1994). They had a
number of advantages over whole heads and florets that were more
cumbersome to handle and produced volatiles that interfered with the
ethylene assay by gas chromatography. Chlorophyll extraction was achieved
by diffusion into methanol overnight, in the dark. These assays provided
simple and accurate means for testing ethylene and chlorophyll levels in
mature broccoli heads.
It was important to establish ethylene production from ‘non-harvested’ buds as a control for ‘harvested buds’. It was clear from the results in Figures
4.3 and 4.4 that as expected, values at 0h post-harvest were very similar to
the non-harvested buds growing in situ. There must have been insufficient
time for a physiological response to harvesting to have taken place. The basal
amount of ethylene produced by non-harvested GDDH33 broccoli buds was
44nl/g bud/h. Basal ethylene is regulated through ACC synthase and aco1
(Pogson et al., 1995) to maintain normal broccoli head physiology in actively
growing plants. Harvesting the head altered the physiology resulting in a
significant increase of ethylene (p<0.005) from 44 to 176nl/g bud/h at 24h.
Pogson et al. (1995) found that the abundance of aco2 transcripts increased
dramatically 24h after harvest, especially in the reproductive organs of the
floral buds. The post-harvest peak of ethylene is probably related to the
increase of aco2 transcripts in these organs, but this would also require a
substantial amount of ACC substrate from ACC Synthase. Tian et al. (1994)
151
and Kasai et al. (1996) observed the same post-harvest ethylene curves as
the GDDH33 control with ‘Green Beauty’, ‘Shogun’, ‘Green Belt’ and ‘Green
Valiant’ broccoli cultivars. There was an increase of ethylene after harvest,
peaking at 24h before declining in these cultivars. The actual values recorded
by Tian et al. (1994) and Kasai et al. (1996) are lower compared to those in
this study since whole florets were used, of which the buds produce the
majority of the ethylene. The post-harvest bud chlorophyll levels also declined
in GDDH33 at the same rate and to the same point as the cultivars used by
Tian et al. (1994) and Kasai et al. (1996). Chlorophyll loss was probably
related to being harvested, as chlorophyll levels in the non-harvested control
did not decline.
The change in post-harvest broccoli bud ethylene production is
associated with the up-regulation of aco2 transcripts within the male and
female reproductive organs (Pogson et al., 1995a). This tissue-specific
expression suggests a role for ethylene in reproduction. In Brassica napus,
ethylene production occurs first at 20 days after pollination (DAP), while a
second greater peak occurs at 35DAP (Rays et al., 2000). In flowers whose
senescence is pollination-dependent or pollination accelerated, pollination is
accompanied by a rapid increase of endogenous ethylene (Hadfield and
Bennett, 1997). The initial pollination induces expression of an ACC synthase
gene in the stigma, which produces ACC that may be oxidised directly to
ethylene or potentially translocated to distal floral organs to be oxidised into
ethylene. Flower organ senescence serves to remobilise nutrients from the
petals and sepals to the developing ovary (Lawton et al., 1989; 1990). It
appears that harvesting forces the immature broccoli buds into premature
152
senescence. The probable role for aco2 is to facilitate remobilisation of
nutrients from surrounding tissues to support the developing ovaries after
pollination. However, aco2 induced by stress after harvest assists the
breakdown of macromolecules into respirable substrate which includes major
losses of sugars, organic acids and proteins (King and Morris, 1994).
The ethylene peak at 24h of 176nl/g bud/h in the untransformed control
was reduced to 116nl/g bud/h in the gus control lines. Although this not a
significant difference (p>0.005) it shows that factors to do with the
transformation and regeneration reduces ethylene production. Henzi et al.
(1999a,b) produced broccoli plants transformed with A. rhizogenes that had
slightly reduced ethylene production compared to the non-transgenic controls.
Henzi et al. (1999a,b) attributed this reduction of ethylene to the rolC locus.
Martin-Tanguy et al. (1993) noted a reduction in ethylene production during
floral development in tobacco plants transformed with rolC (ORF 12 from the
Ri TL-DNA of Agrobacterium rhizogenes), driven by the CaMV 35S virus
promoter. It is possible that the rol genes also caused a reduction of
chlorophyll at 0h between the untransformed 721 g/g bud and 409 g/g bud in
the transformed plants. However, the rate of chlorophyll loss in the
transformed plants was also reduced.
The transgenic lines containing the aco1 antisense construct produced
only 32nl/g bud/h of ethylene at 24h, significantly less than the untransformed
and transformed gus control plants (p<0.005). It shows that the ACO cDNAs
isolated by Pogson et al. (1995a,b) are important in the post-harvest peak of
ethylene of Brassica oleracea. It demonstrates that they can effectively be
down-regulated by antisense constructs transferred through plant
153
transformation. Accompanying the reduction of ethylene was a loss of
chlorophyll that was significantly reduced (p<0.005) compared to the
untransformed control.
Transgenic lines transformed with the aco1 sense construct also reduced
post-harvest production of ethylene to basal levels. The aco1 sense lines
significantly reduced ethylene production compared to the untransformed
control (p<0.005), and did not experience any chlorophyll loss over the 96h.
These lines show that gene silencing may be achieved with both antisense
and sense constructs, and that the CaMV 35S promoter was suitable to
achieve this.
The reduction of post-harvest production of ethylene was achieved with
as much success with the aco2 constructs as with aco1. This suggests that
there has been cross-hybridisation between the transcripts of both genes and
constructs to reduce total amount of ethylene to a basal level. The cDNAs of
aco1 and aco2 share 83% nucleotide identity in their putative translated
regions (Pogson et al., 1995a). aco2 antisense significantly reduced post-
harvest ethylene compared to the untransformed control (p<0.005), and
significantly reduced chlorophyll loss compared to the transformed gus control
(p<0.005). The aco2 sense constructs also manipulated post-harvest
production of ethylene to basal levels, and reduced chlorophyll loss
significantly (p<0.005) compared to the untransformed control. Reduction of
endogenous ethylene to extend shelf-life has not been achieved before with
transformation of broccoli. Experiments with methylcyclopropane an inhibitor
of ethylene action, have shown that storage life of broccoli treated and stored
at 20oC may be extended (Ku and Wills, 1999).
154
The ACC synthase antisense constructs did not decrease total post-
harvest ethylene production as much as the ACO constructs. The acs
constructs shifted the ethylene peak from 24h to 72h (Figure 4.12). The native
ACC synthase is probably normally active 0 - 24h post-harvest as this function
has been lost. It shows that the ACC synthase cDNA obtained from Pogson et
al. (1995) could play a major role in the post-harvest peak of ethylene in
broccoli. It also suggests that there may be another(s) ACC synthase(s) active
between 48 and 72h with a different nucleotide sequence that has not been
silenced. Although there is evidence for the regulation of ACC synthase at the
post-translational level (Jones and Woodson, 1999; Felix et al., 1991, 1994;
Spanu et al., 1994), expression studies indicate that ACC Synthase activity is
most often the result of the increased accumulation of ACC synthase mRNA
transcripts (Kende, 1993; Zarembinski and Theologis, 1994). However,
Pogson et al. (1995b) states that there is no change in expression of this gene
post-harvest, and that the protein might become post-translationally activated.
Kasai et al. (1998) found that the post-harvest increase of ethylene in
senescing broccoli was suppressed by cycloheximide (an inhibitor of
translation), suggesting that it results from the de novo synthesis of the ACC
Synthase enzyme. It would be surprising if transcripts of this gene were not
increased post-harvest as in wounded, ripening tomato (Lycopersicon
esculentum), the ACC synthase protein decays rapidly in vivo with a half-life
of 58 minutes (Kim and Yang, 1992).
The transgenic line transformed with the ACC synthase sense
construct provided more evidence that the acs isolated by Pogson et al.
(1995b) appears to play a role in the post-harvest peak of ethylene. Post-
155
harvest ethylene production at 0h was very high at 245nl/g bud/h, suggesting
high levels in the growing buds. By 24h, this was reduced to only 7nl/g bud/h.
Such a dramatic decrease in ethylene production must either be due to
silencing of the transcripts or inhibition of the ACO or ACS enzymes. If the
CaMV 35S promoter has constitutively produced transcripts in the other
constructs throughout the time period to reduce ethylene production, it is
possible that acs sense transcripts are still produced. It may be hypothesised
that the native ACC synthase is up-regulated post-harvest and together with
the construct produces transcripts above a threshold, leading to co-
suppression of the transcript. As the protein has a high turn over rate (Kim
and Yang, 1992), there would be little ACC substrate for the ACC oxidases to
oxidise to produce ethylene.
ACC synthase is probably the rate-limiting enzyme in broccoli bud
ethylene production. In the acs sense lines, the growing buds produced
245nl/g bud/h ethylene, whereas the buds in the untransformed control
produced only 44nl/g bud/h. If the functional acs sense construct has
produced more mRNA transcript, and more ACC synthase enzyme, without
silencing, this would produce more ACC. If this is the case, the native ACO
present was not limiting the production of ethylene in growing buds with 44nl/g
bud/h, as with more ACC, ACO could catalyse the production of 245nl/g
bud/h. It follows that increasing ACC oxidase activity post-harvest would have
little effect if there was not an increase in the ACC pool, which would probably
be mediated through ACC Synthase(s). Either an up-regulation of the native
acs transcript or post-translational activation of ACS enzyme might increase
the ACC pool. It is possible to propose a model where the rate-limiting step in
156
the biosynthesis of ethylene in Brassica oleracea var. L. italica is controlled by
ACC synthases. ACC synthases appear to play a role in controlling post-
harvest production of ethylene in broccoli buds. It is possible that ACS also
controls normal basal levels of ethylene through ACC production that is
oxidised by ACC oxidases into ethylene. After harvest, the stress could induce
the up-regulation of at least two acs (0-48h and 24-96h) and aco1/aco2 in the
reproductive structures, resulting in an increase in biosynthesis of ethylene.
All of the lines were used to plot ‘total’ ethylene production with the
percentage of chlorophyll loss over the time period. There was a positive
correlation between ethylene production and chlorophyll loss of post-harvest
broccoli buds at 20oC. As ethylene increased, chlorophyll loss increased. This
linear correlation was described by the equation y = 0.2386x – 23.041, where
y = chlorophyll loss and x, the ‘total’ production of ethylene. It was confirmed
by comparing the ratio of the regression and the residual mean squares with
the F-distribution (p< 0.005). ‘Total’ post-harvest production of ethylene
appeared to be an important factor for chlorophyll loss. When the peak varied
between 24 and 72h in different lines, chlorophyll loss was dependent on ‘total’ amount of ethylene, not the timing of maximum production. The
transgenic lines that have down-regulated production of ethylene have also
reduced chlorophyll loss suggesting that chlorophyll loss had a dependence
on ethylene. It was clear that reducing post-harvest ethylene production to the
basal level produced in non-harvested buds, prevented chlorophyll loss.
Although ethylene production and post-harvest decay have been significantly
positively correlated in melon fruits (Zheng and Wolff, 2000), this has not yet
been determined in broccoli. Chlorophyll loss is an important factor because
157
yellowing is the last stage of senescence preceded by the breakdown of
sugars, organic acids and proteins (King and Morris, 1994). It demonstrates
that the shelf-life of a very perishable vegetable may be increased at least up
to 3 days at 20oC by reducing post-harvest ethylene. This should be of benefit
to retailers and consumers preserving the nutritional quality of the product for
a longer time period.
In the experiments described here, the data were collected from
primary transformants, which means that they are effectively heterozygous for
the constructs. In chapter 3, seeds were harvested from 11 of the T0 lines,
and 67 seeds were sown from the 28/00 aco1A 3 line. It showed that the TL-
DNA, carrying the rol genes may be removed by segregation. As the rolC
locus appears to reduce endogenous ethylene production (Martin-Tanguy et
al., 1993) it would therefore be more useful to analyse the T1 generation for
the effects of the constructs on ethylene production and bud chlorophyll loss.
The T1 generation may also be homozygous for construct T-DNA inserts and
this may show more extreme effects. It also allows the opportunity for
crossing the Tx lines containing different constructs, such as an acs antisense
line with an aco1/aco2 antisense. A Tx line with stacked genes may produce
even less ethylene post-harvest, and possibly produce a phenotype with a
longer shelf-life. There is a possibility that construct copy number and, the
effect of their position within the genome may lead to differences of mRNA
expression. There was no evidence in chapter 3 that copy number of the TL-
DNA related to the severity hairy root phenotype, and the T0 line with greatest
reduction of post-harvest ethylene production (3/00 aco2A 2) appeared only to
have one insert. It is possible that the ‘position effect’ of construct within the
158
plant genome had a greater influence over mRNA expression, and effects on
the plant phenotype.
This research has showed that there was a relationship between post-
harvest production of ethylene and chlorophyll loss in broccoli heads. aco1,
aco2 and acs appear to play major roles in the post-harvest production of
ethylene, although there may be more, equally important genes. The native
genes are probably up-regulated following the stress caused by harvesting,
forcing the buds into premature senescence that would usually be required to
support the developing ovary after pollination. An assay was set up to
measure ethylene and chlorophyll levels using broccoli buds that proved to be
reliable and efficient. The whole process shows that genes can be targeted
and via a transformation system be down- or up-regulated to study
characteristics in vivo, to either understand the genetic control behind
physiological responses or introduce beneficial traits to complement breeding
programmes. In this case, utilising the aco1, aco2 and acs cDNAs isolated
from Brassica oleracea L. var. italica by Pogson et al. (1995a,b) has helped
understand the genetic control behind the senescence of broccoli buds and
may enable the manipulation of the post-harvest physiology of broccoli.
159
5.0 Identifying Brassica oleracea
ACC oxidases
160
5.1 Introduction
Ethylene biosynthesis occurs via the enzymes ACC Synthase (ACS)
and ACC Oxidase (ACO), which catalyse the conversions of S-
adenosylmethionine (SAM) to 1-aminocyclopropane-1-carboxylic acid (ACC)
and ACC to ethylene, respectively (Kende, 1993). In broccoli, ethylene
appears to play an important role in sepal yellowing after harvest, since
chlorophyll loss is associated with an increase in ethylene production (Tian et
al., 1994). The post-harvest increase of ethylene synthesis is associated with
an increase of ACC oxidase 1 and ACC oxidase 2 mRNA transcripts in the
broccoli florets (Pogson et al., 1995a).
ACC oxidases have been isolated and characterised from multigene
families from three species including tomato (3 genes), petunia (4 genes) and
melon (3 genes) (Barry et al., 1996; Tang et al., 1993; Lasserre et al., 1996).
ACC oxidases show very different expression patterns in various tissues at
different stages of development and in response to specific stimuli (Barry et
al., 1996). Pogson et al. (1995a) isolated two cDNAs encoding ACC oxidases,
aco1 (1237bp) and aco2 (1232bp), from broccoli. aco1 and aco2 share 83%
nucleotide identity in their putative translated regions, 48% in the untranslated
regions, and share 86% amino acid identity (Pogson et al., 1995a). Southern
analysis of the aco1 and aco2 at high stringency gave distinct hybridisation
patterns, confirming that they encoded by different genes (Pogson et al.,
1995a). Low stringency analysis revealed only one additional hybridising band
not accounted for by either aco1 or aco2, indicating that broccoli may contain
at least one other closely related gene (Pogson et al., 1995). aco1 and aco2
appear to be important genes in the post-harvest senescence of broccoli as
161
their protein products catalyse the last step in the ethylene biosynthetic
pathway. If ACO1 and ACO2 do play a major role in post-harvest senescence
of broccoli it would be useful to understand regulation of the gene family, and
characterise different alleles as these may produce phenotypes with reduced
post-harvest ethylene production. Allelic variation could be determined by
genetic markers and assist efforts to lengthen post-harvest shelf-life of
broccoli.
Brassica oleracea is considered a diploid species (2n=18) along with B.
nigra (2n=16) and B. rapa (2n=20). Pairwise combinations of the diploid
species B. rapa (syn. Campestris)(AA), B. nigra (BB) and B. oleracea (CC)
have resulted in the amphidiploid species B. juncea (AABB), B. napus (AACC)
and B. carinata (BBCC) (U. 1935). Many genetic maps have been assembled
for Brassica species (reviewed by Paterson et al., 2000). Sebastion et al.,
(2000) constructed an integrated AFLP and RFLP Brassica oleracea linkage
map from F1 crosses of Brassica oleracea doubled haploid populations:
Chinese Kale X Calabrese (var. alboglabra X var. italica) and cauliflower X
Brussels sprouts (var. botrytis X var. gemmifera) (Sebastion et al., 2000).
Integration of the map was based on 105 common loci including RFLPs and
AFLPs to provide a consensus map of 547 markers mapping to nine linkage
groups. It is thought that the integrated B. oleracea genetic map will facilitate
the mapping of loci across the populations (Sebastion et al., 2000).
Bacterial artificial chromosomes (BACs) are large genomic DNA
clones, ranging from 100 to 200kb, used for map based cloning (Jackson et
al., 1999). A BAC library has been constructed for Brassica oleracea from the
162
doubled haploid alboglaba12 (G. King, pers. comm. HRI-Wellesbourne), and
can be used for cloning genes.
The aim of this section was to: a) confirm whether aco1 and aco2 are
different genes, and if so how many copies are there of each gene in the
Brassica oleracea genome; b) try to determine whether there are any other
aco genes in the Brassica oleracea genome; c) locate their genetic map
positions with the integrated linkage map from (Sebastion et al., 2000) and; d)
to determine how similar the Brassica oleracea var. italica aco genes are to
other Brassica aco genes.
163
5.2 Materials and Methods
5.2.1 PCR
Primers were designed to produce fragments from B. oleracea aco1
and aco2 cDNAs and genes (see Table 5.1). All oligonucleotides were
synthesised by MWG-Biotech (MWG-BIOTECH AG Anzinger Str.7 D-85560
Ebersberg) (Table 5.1) and used in standard PCR conditions (Tables 5.2 and
5.3). PCR components including Taq polymerase, PCR buffer and dNTPs
were obtained from GIBCO-Life Technologies, Uxbridge, UK.
Table 5.1 Oligonucleotide sequences used for primers to determine gene polymorphisms and produce DNA probes .
Primer
Name
Sequence (5’- 3’) Tm (a) (oC)
2 GCT TTG GTC GAC GAT GCT TG 62
3 AGT CTC TAC GGC TGC TGT TG 62
4 CGT CTC GAC TGT GGC CAC CG 68
5 TAG GAC AAG CTG GAT AGT TGC T 61
6 CTC ACA CTG ATG CAG GAG GC 65
7 CCT GAT ATG TCT GAT GAA TAC CG 61
8 CCT CAG CAA GAT TCT CCA ATC 61
9 TGG TCG ACG ATG CTT GTC A 60
10 GTC CAT AAG ATC ATA TGG TAT TCC 59
aco1 3’UTR F GAA GAA TTC TAA TGC AGT TAC AG 57
aco1 3’UTR R CGC AAG CTT AGA ACA GTT ATC ATA T 59
aco2 3’UTR F GAA GAA TTC TGC AGC AAC CAC CGA TTT G 66
aco2 3’UTR R GCG GAA GCT TTA AAA TTC ATA TTT CCA ATA C 62
aco2 R TCA TAT TTC CAA TAC AAT TAT TGC 54
aco2 F CTA ACA TCA TCA ACA ACA TTA AG 56
(a) Tm of the primers was calculated using the expression:
Tm = 81.5 + -16.6 + (41 x (#G + #C/length)) -(500/length)
164
Table 5.2 PCR reaction components. To make a total 25 l for each reaction, 5 l of template DNA (100pmol/ l) and 5 l of primers (set at 4pmol/ l) were added. Primers and template DNA were suspended in autoclaved R.O water.
x1 ( l) 10x Buffer 2.5 50mM MgCl2 0.75 dNTPs 25mM 0.2 H2O 11.42 Taq 0.13 Total 15
The PCR program in Table 5.3 was used and the annealing
temperature adjusted 3oC below the Tm of primer pairs (Table 5.1). The PCR
reactions were performed on a Hybaid Omnigene PCR machine.
Table 5.3 PCR program. Stages 1 and 3 were cycled once. Stage 2 was cycled thirty five times. Stage 1. 1) 93 C/ 2 min 2) 53 C/ 1 min
Stage 2. 1) 72 C/ 1 min 2) 93 C/ 30 sec 3) 53 C/ 1 min
Stage 3. 1) 72 C/ 10min.
The location of the primers (Table 5.1) on the ACC oxidase 1 and ACC
oxidase 2 cDNAs are shown Figures 5.1 and 5.2. Figures 5.1 and 5.2 are
included as a reference for sequences that are amplified by PCR.
165
Exon1 Intron1 Exon2 Intron2 Exon3 Intron3 Exon4
Exon1 Intron1 Exon2 Intron2 Exon3 Intron3 Exon4
2,9 7
5’
aco1 3’UTR F
3’
10 8 5 aco1 3’UTR R
Figure 5.1 Schematic diagram of the predicted Brassica oleracea L. var. italica ACC oxidase 1 gene. The arrows and numbers represent forward and reverse primer locations (see Table 5.1).
aco2 F
5’
2,9 7 aco2 3’UTR F
3’
10 8 5 aco2 3’UTR R aco2 R
Figure 5.2 Schematic diagram of the predicted Brassica oleracea L. var. italica ACC oxidase 2 gene. The arrows and numbers represent forward and reverse primer locations (see Table 5.1).
5.2.2 Bacterial Artificial Chromosomes (BACs)
E.coli carrying the Brassica oleracea physical map BACs were cultured
in 10mls LB with 10 l (50mg/ml) chloramphenicol, in sterile 50ml centrifuge
tubes, overnight at 37oC. Three millilitres of cells were added to 1.5ml
microfuge tubes, centrifuged at 11600g for 1min, and the supernatant
discarded. 0.3ml Qiagen P2 lysis buffer was added to the pellet and left for
5min at room temperature, then 0.3ml of pre-chilled Qiagen P3 neutralisation
buffer was added and left for 5min on ice (Qiagen Ltd., Boundary Road,
Crawley, West Sussex). The tubes were centrifuged at 11600g for 10min, and
the 450 l supernatant was added to 1125 l 100% ethanol (analaR grade,
166
BDH) and 32 l 5M sodium acetate. The tubes were centrifuged at 11600g for 30min, and the supernatant was removed allowing the pellet to air dry for
5min at room temperature. The pellet was resuspended in 100 l sterile,
purified H2O. The DNA was quantified on a 1% (w/v) agarose gel.
An EcoRI endonuclease restriction digest was carried out at 37oC overnight with the BAC DNA in the following reaction: 17 l DNA (85ng); 2 l
reaction buffer 3; 0.5 l EcoRI enzyme (GIBCO-Life Technologies); 1 l
spermidine (10mg/ml); 0.1 l RNase (10mg/ml), to make a final volume of
20.6 l. 5.2.3 Southern Blots and fliters
The restriction endonuclease digestion products from section 5.2.2
were loaded onto a 0.8% (w/v) agarose gel and separated by gel
electrophoresis for 15h at 50V. They were stained in 0.33 g/ l ethidium
bromide, photographed, and then de-stained in R.O. water. The blotting was
performed by alkaline lysis onto Hybond N+ filters (Amersham) according to
Sambrook et al. (1989).
The Brassica oleracea BAC library (BoB) filters, and the AG and NG
mapping filters were obtained from Graham King (HRI-Wellesbourne). The
mapping filters are derived from the integrated map of B. oleracea based on
the segregation data of F1-derived doubled-haploid mapping populations
(Sebastion et al., 2000).
167
5.2.4 Hybridisation between N+ filters and 32p dCTP labelled DNA probes
The Hybond N+ filters were placed onto nylon mesh that had been
wetted in 2x SSC (saline sodium citrate), wrapped up tightly and placed in a
Hybaid hybridisation bottle. Fifty millilitres of hybridisation buffer (0.5M Na
phosphate pH 7.2; 7.0% sodium dodecyl sulphate [SDS]; 10mM EDTA, pH
8.0 and; 100 g/ml single stranded salmon sperm DNA) was added to the
tubes at 55oC. The tubes were placed in a Hybaid rotisserie oven at 55oC and
left to pre-hybridise for 3h. DNA probes were by produced by PCR with
primers aco1 3’UTR F, aco1 3’UTR R, aco2 3’UTR F, aco2 3’UTR R, aco2 R, aco2
F, set at the conditions of Tables 5.2 and 5.1, and purified with the QIAquick
PCR Purification Kit (Qiagen Ltd., Boundary Road, Crawley, West Sussex).
The DNA probes were labelled with 32P dCTP according to the rediprime II
random prime labelling system kit (Amersham Pharmacia Biotech, Bucks.
UK). The labelled DNA was separated from the other components by a
sepharose column and counted with a scintillation counter. The 50mls of
hybridisation buffer were replaced with 10mls fresh buffer, and 2 x 107 cpm (in
2x106/ml) labelled probe and left overnight at 55oC. The blots were washed at
low stringency with 10 x SSC and 1% SDS four times at 55oC. The filters were
wrapped in Saranwap (Dow Chemicals) and placed into X-ray cassettes with
XAR5 film (Kodak) under a safelight. The cassettes were placed at –80oC for
48h, and then the film was developed. 5.2.5 Genetic Mapping
The B. oleracea mapping filters obtained from Graham King (HRI-
Wellesbourne) were hybridised to the 32P dCTP labelled aco1 3’UTR and
168
aco2 3’UTR DNA probes. These experiments produced autoradiographs that
were scored by James Higgins for segregation of progeny from the parents.
The data was analysed in conjunction Dr Graham Teakle (HRI-Wellesbourne)
with the computer software JOINMAP 2.0 (Stam et al., 1995), to calculate a
map position.
5.2.6 Cloning and sequencing PCR fragments
DNA fragments were amplified by PCR from BACs 3(10 J 22), 4(11 K 12), 12(43 P 13), 20(62 O 07), with primers 2 and 5 (Table 5.1). DNA products
from 3 reactions of 4 BACs (12) were ligated into the pCRII-TOPO cloning
vector, with the TOPO TA cloning kit (Invitrogen). The transformation of the
ligated plasmid was carried out according to the protocol supplied with the
DH5 library efficient competent cells (E. coli DH5 , genotype: F-
80dlacZ M15 (lacZYA-argF)U169 deoR recA1 endA1 hsdR17(rk-, mk
+) phoA supE44 -thi-1 gyrA96 relA1) (GIBCO-Life Technologies, Uxbridge,
UK.). Cells were cultured in LB media and selected for with 25mg/l
kanamycin. Single colonies were checked for presence of the insert with a
colony PCR method taken from the pMOSBlue T-vector kit (Amersham
International plc, Little Chalfont, Buckinghamshire, UK), using primers 2 and
5. The plasmids from positive colonies were extracted with Qiagen miniprep
(tip20) kits (Qiagen Ltd., Boundary Road, Crawley, West Sussex) and the
PCR fragment inserts were sequenced at Birmingham University.
DNA was extracted from agarose gels with the QiaexII kit (Qiagen,
Boudary Court, Crawley, W. Sussex) and sequenced by SEQLAB, Gottingen.
169
5.2.7 Analyses of sequences
Nucleotide sequences were identified and compared with BLAST
searches (http://www.ncbi.nlm.nih.gov/blast/), and the similarities and possible
phylogeny determined by MegAlign computer software of DNAStar
(Lasergene).
170
5.3 Results
5.3.1 Identifying Brassica oleracea BACs containing ACC oxidase genes
5.3.1.1 Isolating BACs containing ACC oxidases
In Figure 5.3 (a,b) the autoradiographs show ‘hits' where the B.
oleracea var. italica full length ACC oxidase 2 cDNA probe has hybridised to
E. coli colonies carrying BACs with homologous regions of DNA. The
hybridisation was carried out at 55oC, and washed at low stringency, so the
aco2 cDNA sequence should hybridise to both aco1 and aco2 and any other
closely related genes, and not be washed off. The BAC codes for the ‘hits’ are
shown in Table 5.4.
(a) (b)
Figure 5.3 Autoradiograph showing hybridisation between the Brassica oleracea BAC library filter 1, (a) and filter 2, (b) and the Brassica oleracea var. italica ACC oxidase 2 full length cDNA (Pogson et al., 1995a).
171
Table 5.4 The codes of the B. oleracea BAC colonies that hybridised to the B. oleracea ACC oxidase 2 full length cDNA probe.
Hit No.
Bob BAC code
Hit No.
Bob BAC code
1 02 D 02 12 43 P 13 2 10 I 20 13 46 E 09 3 10 J 22 14 50 C 13 4 11 K 12 15 50 O 10 5 12 N 02 16 51 M 19 6 23 I 16 17 51 N 21 7 29 D 02 18 60 E 16 8 29 O 22 19 61 I 11 9 34 O 03 20 62 O 07 10 35 E 16 21 62 I 13 11 37 K 18 22 65 I 13
5.3.1.2 Identifying ACC oxidases from BACs isolated
In Figure 5.4, the aco1 3’ UTR probe hybridised to BAC hit numbers
11, 12, 13 and 15. In Figure 5.5, the aco2 3’ UTR probe hybridised to BAC
colonies 9, 10, 16, 20, and 22. There appears to be specific binding of the
aco1/aco2 probes to the aco1/aco2 genes and no cross-hybridisation. The
283bp aco1 and 252bp aco2 3’UTR probes share only 44% nucleotide
identity, spread evenly along the sequences. There was specific hybridisation
of the aco1 probe to BAC colonies 11, 12 and 15 (Figure 5.4) and the aco2
probe to BAC 20 (Figure 5.5). If single bands suggest single gene copies,
these data suggest that BACs 11, 12, 15 contain one single aco1 gene and
BAC 20 contains one single aco2 gene. This leaves 18 BACs unaccounted for
which contain sequences that originally hybridised to the whole length B.
oleracea aco2 cDNA sequence.
172
bp
12000
5000
2000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Figure 5.4 Autoradiograph showing hybridisation between the B. oleracea ACC oxidase 1 cDNA (Pogson et al., 1995a) 3’ UTR 283bp probe and the EcoRI RFLP BAC filter. In lanes 1 and 24, the 1kb Plus DNA Ladder (GIBCO-Life Tehnologies) and
in lanes 2-23 BAC clone hit numbers left-right: 1, 5, 11 ,12, 13, 14, 15, 19, 21, 2, 3, 4, 6, 7, 8, 9, 10, 16, 17, 18, 20, 22 (see Table 5.4).
173
bp
12000
5000
2000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Figure 5.5 Autoradiograph hybridisation between the B. oleracea ACC oxidase 2 cDNA (Pogson et al., 1995a) 3’ UTR 252bp probe and the EcoRI
RFLP BAC filter. In lanes 1 and 24, the 1kb Plus DNA Ladder (GIBCO-Life Tehnologies) and in lanes 2-23 BAC clone hit numbers left-right: 1, 5, 11 ,12, 13, 14, 15, 19, 21, 2, 3, 4, 6, 7, 8, 9, 10, 16, 17, 18, 20, 22 (see Table 5.4).
174
In Figure 5.6, the aco1 specific primers amplified the same size
fragment (1300-1400bp) from all 22 BACs. The BAC sequences must have
had high homology to the primers for amplification and producing the same
size fragment suggests that all of the BACs contain ACC oxidase sequences.
The positive control from the cDNA (889bp) was smaller than the product from
the BACs that were 1300-1400bp. This suggests that all of the BACs contain
ACC oxidases with introns that amount to 400-500bp in the region between
primers 2 and 5. These data suggest that either all BACs contain an aco1
sequence or that these primers do not distinguish between aco1 and aco2
sequences (or any other closely related sequences).
The aco2 specific primers (2 and 4) amplified fragments in Figure 5.6,
from only 5 BAC hits, 9, 10, 16, 20 and 22. These were the same BAC clones
that the aco2 3’ UTR probe hybridised to on the EcoRI BAC RFLP filter. It
appears that these colonies contain aco2 genes, thus leaving 17 BACs that
need characterising.
PCR products from the amplification of BAC hit numbers 3, 4, 12 and 20 with primers 2 and 5 were ligated into a vector for sequencing. Three
individual colonies from each transformation were chosen for sequencing in
forward (primer 2) and reverse (primer 5) orientations. Sequencing BACs 3
and 4 produced two separate sequences that aligned to both aco1 and aco2
cDNA sequences in a BLAST search (http://www.ncbi.nlm.nih.gov/blast/).
BAC 12 contained a single aco1 sequence, and BAC 20 contained an aco2
sequence, when entered into a BLAST search
(http://www.ncbi.nlm.nih.gov/blast/). The aligned consensus sequences of
aco1 and aco2 are shown in Figures 5.7 and 5.8. It is clear in Figures 5.7 and
175
5.8 that aco1 and aco2 are different genes. Intron 1 in aco1 is much smaller
(97bp) than the aco2 intron 1 (248bp) and they share only 18% nucleotide
identity. The intron 2 of aco1 is much larger (312bp), than that of aco2
(108bp) and they share only 21% nucleotide identity. There is no homology
shared between introns of Brassica oleracea aco1 and aco2.
(a)
bp
2000 1200 800 400
(b)
bp
2000 1200 800 400
(c)
bp 2000 1200 800 400
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Figure 5.6 Electrophoretogram showing the products of a PCR between the BAC DNA and aco1 specific primers (2 and 3) and aco2 specific primers (2 and 4). Lanes 1 and 2 contain the Step ladder and Low DNA Mass ladder, respectively (GIBCO-Life Technologies). The 7th lane in row (a) is the positive control from the cDNA of aco1. Lanes 8 –20 (a), and 3-11 (b) are the PCR products from primers 2 and 3 on all of the BACs. The 12th lane (b) is the positive control from the cDNA of aco2 for primers 2 and 4. Lanes13-20 (b) and 3-15 (c) are the PCR products from primers 2 and 4 from the BAC hits 1- 22.
176
GTACATATATATTTTTTCTTCATAAAATCAACTTTAAATCATATGTTATG GTA
GTA
AC CA
CA --
ACO1 :
ACO2 :
* 20 * 40 * 60
------------------------------------GTACAAGCATATATGTGATTATATCTAG--
GTTATCGTCTCTCTACTCGTCCTTTTTTTGTTTACAATCAAAGAAAATATAATAATATAACNGAGC
T AAG A ATAT A TATA C
: 28
: 66
ACO1 :
ACO2 :
* 80 * 100 * 120 *
-----------------CTTTTTTGAGTTTGTGTACTTAATTG-----------------------
NTTAGATGTTCTTTGCTCAAACNTTTGTTGGTGTATAAAATATGAAAAACTTKGAWTATATATATA
C T GTT GTGTA AAT
: 54
: 132
ACO1 :
ACO2 :
140 * 160 * 180 * 2
-----------------------------------GTAATGTGGATCTTTTTGTTTGGTGGTTAAC
TATATATATATATATATATGTTATTATTRAGCTATGAAATGACGCAATARATATTATATACCTAGC
G AATG G T T TT T TA C
: 85
: 198
ACO1 :
ACO2 :
00 * 220 * 240 * 260
TTG---------ATTTTCCAG---------------------------------------------
TTCGAARACATCATTTTCCTTTCATAATTTTTTAAAAAATGGGTCGTTATAACTCTTGATATTTGT
TT ATTTTCC
: 97
: 264
ACO1 :
ACO2 :
*
--------
AATATCAG
: -
: 272
Figure 5.7 Brassica oleracea ACC oxidases 1 and 2 intron 1 aligned to determine possible homology. Red boxes with white letters= consensus; Blue boxes with yellow letters= no consensus. A= adenine, T=thymine, C=cytosine, G= guanine.
ACO1 :
* 20 * 40 * 60
AAAAATATC
: 66
ACO2 : -------------------------------------------------- CACACAAAC
GTA CA A A A C
: 14
ACO1 :
ACO2 :
* 80 * 100 * 120 *
ATATGTTATATCCCCTTTAAAAGGGCCAC-TCTGCCACTTT-TACCTATATTAAAAA---
GATTTT ATTTCTT---TCACCCTTAAAT----
CACATCTCCAACTATCTTATTAAAATAAACCATTGATTTG AT T TT TC CC TTAAA
CAC TCT C ACT T T TA A TAAA GATTT
: 127
: 73
ACO1 :
ACO2 :
140 * 160 * 180 * 2
TGTGATATTTTATTTCTAAACAAAATAACTATACTTTGTTAGTTAGTAAAAACAGTTTTAAGGAAT
CTTGCCATGTT-TCTTTAAAACACAGAACGG--CTGAG---------------------------- TG AT TT T T TAAA A A AAC CT G
: 193
: 108
ACO1 :
ACO2 :
00 * 220 * 240 * 260
TTGTTTCACTTTAGAACCTCTAATCCTTTTTGTGTAATGAAAATAAAGTTGAGAAGAAACGTCTAA
------------------------------------------------------------------
: 259
: -
ACO1 :
ACO2 :
* 280 * 300 *
AAATTTAACACACTTATTTGAAAGAGGCATACTGAAATGTTTTTATTTTGCA
G
-----------------------------------------------------
: 312
: -
Figure 5.8 Brassica oleracea ACC oxidases 1 and 2 intron 2 aligned to determine possible homology. Red boxes with white letters= consensus; Blue boxes with yellow letters= no consensus. A= adenine, T=thymine, C=cytosine, G= guanine.
177
Primers were designed to amplify intron 1 (9 and 10) and intron 2 (5
and 7) for both aco1 and aco2 (see Figures 5.1 and 5.2). Primers 5 and 7
should amplify a 500bp DNA fragment by PCR with an aco1 template and a
280bp fragment with aco2. Primers 9 and 10 should amplify a 130bp DNA
fragment with an aco1 template and a 300bp with aco2. These fragments
could be separated on a 1% (w/v) agarose gel by electrophoresis to show
polymorphism between aco1 and aco2. In Figure 5.9, the BACs that had been
sequenced (3, 4, 12 and 20) were tested with primers 9 and 10, and 2 and 5
for evidence of aco1 or aco2 introns. Intron 2 was amplified with primers 5
and 7 and produced the lower 280bp fragment from aco2 and the larger
(500bp) fragment of aco1. The PCR showed that BACs 3 and 4 contained
both aco1 and aco2, whereas BAC 12 contained only aco1 and BAC 20, only
aco2 sequences.
bp
1353 1087 872 603 310
1 2 3 4 5 6 7 8
Figure 5.9 Electrophoretogram showing the PCR products from primers 5 and 7 on BAC’s 3, 4, 12 and 20. Lane 1, X174 RF DNA/HaeIII fragments; lane 2, BAC 3; lanes 3 and 4 BAC 4; lanes 5 and 6; BAC 12; lanes 7 and 8, BAC 20.
178
In Figure 5.10, primers 5 and 7, and primers 9 and 10 were used to
amplify fragments from all of the BACs. BAC 12 has evidence for only having
an aco1 gene. The other 21 BACs have evidence for having both aco1 and
aco2. The PCR fragments A-D in Figure 5.10 were extracted from the gel and
sequenced. They were the same sequences as the aco1 and aco2 introns 1
and 2.
(a) bp
1353 1078 872 603
310
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
A B
(b)
bp
1353 1078 872 603
C 310 D
(c)
bp 1353 1078 872 603
310
Figure 5.10 Electrophoretogram showing PCR products from all BACs with primers 5 and 7, and, 9 and 10. Lane 1 for rows (a-c) is the DNA ladder of X174 RF DNA/ HaeIII fragments. In row (a), lanes 2-20 show PCR products from primers 5 and 7 and BACs 1- 19. In row (b), lanes 2-4 show PCR products from primers 5 and 7 and BACs 20-22. Lanes 5-20 show PCR products from primers 9 and 10 with BACs 1- 16. In row (c), lanes 2-7 show PCR products from primers 9 and 10 with BACs 17-22. Arrows: A= aco1 intron 2; B=aco2 intron 2; C= aco2 intron 1; D=aco1 intron1.
179
Figure 5.11 shows the autoradiograph produced from the hybridisation
between the BAC EcoRI RFLP filter and the aco1 intron 2 probe, amplified by
PCR from BAC 12 with primers 5 and 7. The intron probe hybridised to the
same locations as the aco1 3’ UTR probe in Figure 5.4. The hybridisation to
BACs 11 and 12 at 12000 and 5000bp, respectively, suggest that the B.
oleracea aco1 gene is located on these two BACs. It also suggests that
originally, the full length cDNA aco2 probe hybridised to this sequence at
55oC, with low stringency washings.
bp
12000
5000
1 2 3 4 5 6
Figure 5.11 Autoradiograph showing the hybridisation between the Brassica oleracea BAC EcoRI RFLP filter and the Brassica oleracea aco1 intron 2 probe. In lane 1 the 1kb Plus DNA Ladder (GIBCO-Life Tehnologies) and in lanes 2- 6 BAC clone hit numbers left-right: 1, 5, 11 ,12, 13.
180
5.3.2 Genetic Mapping of Brassica oleracea aco1 and aco2 genes
Figure 5.12, shows the autoradiograph produced from the hybridisation
between the aco1 3’UTR probe and the EcoRI digested DNA from the cross
between A12 and GD33 (AG population). There are clearly 2 major bands,
suggesting that there are two copies of the aco1 gene that have hybridised to
the probe. However, the lines have the same binding pattern so there is no
polymorphism to score for segregation of the progeny.
Figure 5.13, shows the autoradiograph produced from the hybridisation
between the aco1 3’UTR probe and the EcoRI digested DNA from the cross
between Nedcha and Gower (NG population). In this population there were
also two major bands containing sequences with high homology to the aco1
3’UTR probe. There was segregation of the progeny in the DH lines of the NG population so it was possible to score the autoradiograph.
Figure 5.14, shows the autoradiograph produced from the hybridisation
between the aco2 3’UTR probe and the EcoRI digested DNA from the cross
between A12 and GD33 (AG population). There were two major bands where
the probe had hybridised to, and segregation of the progeny that could be
scored.
Figure 5.15, shows the autoradiograph produced from the hybridisation
between the aco2 3’UTR probe and the EcoRI digested DNA from the cross
between Nedcha and Gower (NG population). There were also two major
bands in this population containing sequences with high homology to the aco2
3’UTR probe and segregation of the progeny.
181
(a)
(b)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
Figure 5.12 Autoradiograph showing the hybridisation between the B. oleracea ACC oxidase1 cDNA 3’ UTR 283bp probe and the AG mapping filter. In row (a) the parents (Alboglabra12 and GD33) are in lanes 3-5 and, in row (b) the parents (Alboglabra12 and GD33) are in lanes 3 and 4. In row (a), lanes 6-33, and in row (b) lanes 5-32 contain the progeny of the cross.
182
(a)
(b)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
Figure 5.13 Autoradiograph showing the hybridisation between the B. oleracea ACC oxidase1 cDNA 3’ UTR 283bp probe and the NG mapping filter. In row (a) the parents (Nedcha and Gower) are in lanes 3-5 and, in row (b) the parents (Nedcha and Gower) are in lanes 3 and 4. In row (a), lanes 6-33, and in row (b) lanes 5-32 contain the progeny of the cross.
183
(a)
(b)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
Figure 5.14 Autoradiograph showing the hybridisation between the B. oleracea ACC oxidase2 cDNA 3’ UTR 252bp probe and the AG mapping filter. In row (a) the parents (Alboglabra12 and GD33) are in lanes 3-5 and, in row (b) the parents (Alboglabra12 and GD33) are in lanes 3 and 4. In row (a), lanes 6-33, and in row (b) lanes 5-32 contain the progeny of the cross.
184
(a)
(b)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
Figure 5.15 Autoradiograph showing the hybridisation between the B. oleracea ACC oxidase2 cDNA 3’ UTR 252bp probe and the NG mapping filter. In row (a) the parents (Nedcha and Gower) are in lanes 3-5 and, in row (b) the parents (Nedcha and Gower) are in lanes 3 and 4. In row (a), lanes 6-33, and in row (b) lanes 5-32 contain the progeny of the cross.
185
The autoradiographs in Figures 5.12-5.15 were scored for segregation
of aco1 and aco2 in the progeny from the parent crosses. The data was
entered into a spreadsheet and analysed in conjunction with Dr Graham
Teakle at HRI-Wellesbourne, with the computer sotware JOINMAP 2.0 (Stam
et al., 1995). JOINMAP 2.0 (Stam et al., 1995) predicted linkage of aco1 and
aco2 genes in Brassica oleracea as shown in Figure 5.15. Of the predicted 2
genes x 2 loci, only one locus from each gene was calculated due to the lack
of polymorphism of the EcoRI mapping filters.
O1 O8
AC-CAR02
pN202E1
BoRGL-IIIa
AC-CACE10 pO87R2
RM4-CAPS-1
pN59E1NM
labi8E2
LEW21
38.1 39
42.1 46.9 50.1 52.1
BoACO1
42.9
pN173E1 pO92E2
pO152E2
pN23J1
pN21E1
84.4
85.6
87.2 92.1
BoACO2
89.3
Figure 5.16 The linkage map positions of aco1 and aco2 in Brassica oleracea for the loci that could be scored by EcoRI RFLP polymorphism. Genetic markers are shown on the left-hand side and map distances (cM) on the right-hand side of the linkage groups.
186
5.3.3 Analysis of nucleotide sequences of closely related genes to Brassica
oleracea aco1 and aco2
The B. oleracea ACO1 cDNA sequence is very similar to the Brassica
juncea ACO mRNA (96.7% nucleotide identity) (emb/z11750/BJEFEMR). It is
also very similar to the Brassica napus (gene bank ref:
gb/L27664/BNAACCOX1) mRNA sequence with 70.8% and Brassica rapa
70% nucleotide identity. The Brassica napus ACC oxidase is the only
sequence to have homology with the B. oleracea aco1 intron 1, but most
Brassica aco sequences have been obtained from cDNAs and do not show
the introns.
The cDNA sequence of B. oleracea aco2 is most like the Brassica rapa
aco mRNA with 80.9% nucleotide identity. It is also very simlar to the B.
napus sequence (gb/L27664/BNAACCOX1) with 77% identity. The B. napus
Introns 1 and 2 of this sequence share 93% identity with the aco2 introns
sequenced in these experiments. There is also high similarity between the B.
oleracea aco2 and the aco cDNA from B. juncea with a nucleotide identitie of
70.1 %.
These sequence analyses show high conservation of aco1 and aco2
within the Brassica family.
187
5.4 Discussion
The aims of this chapter were to determine how many aco genes exist,
their copy number and where they mapped to in the Brassica oleracea
genome. The first stage was to identify aco sequences, and the second to
map these sequences.
The full length B. oleracea var. italica aco2 cDNA sequence (Pogson et
al., 1995) was hybridised to the Brassica oleracea BAC (BoB) library and
washed at low stringency. This produced twenty-two BAC colonies that
hybridised to the cDNA probe. The BAC EcoRI RFLP filter was concurrently
hybridised to the aco1 3’UTR and aco2 3’UTR sequence specific probes. The
aco1 3’UTR probe hybridised to BAC hit numbers 11, 12, 13 and 15 and the
aco2 3’UTR probe hybridised to BAC hit numbers 9, 10, 16, 20 and 22.
However, specific primers for aco1 amplified fragments from all of the BACs.
Specific primers for aco2 only amplified fragments from the BACs that the
aco2 3’UTR probe had hybridised to. It was difficult to interpret this data, so
four out of the 22 BACs were sequenced between primers 2 and 5 of the ACC
oxidase sequence. Exon sequences were aligned with the either aco1 or aco2
cDNA sequences from Pogson et al. (1995a). The exon sequences of BAC 12
matched only aco1 and BAC 20 only aco2. BACs 3 and 4, which had not
hybridised to either of the 3’UTR probes, appeared to have both aco1 and
aco2 sequences. Sequences matched up to either aco1 or aco2, so BACs
that hybridised to the full length aco2 cDNA probe, but not to the 3’UTR
probes may contain aco1 and/or aco2 genes.
The sequences for introns 1 and 2 from B. oleracea aco1 and aco2 were obtained. They showed that aco1 and aco2 were two different genes,
188
and they did not share homology between the introns. The aco1 intron 1
(312bp) sequence was amplified and hybridised to the BAC EcoRI RFLP
filter. The aco1 intron 1 sequence hybridised to the same places as the aco1
3’UTR probe, suggesting that they were hybridising to the same fragment.
The polymorphism between intron sizes was used to determine the identities
of BAC sequences that had originally hybridised to the B. oleracea aco2 full
length cDNA probe. The results appeared to show that all of the BACs
contained an aco1 and aco2 gene, except BAC 12, which only had evidence
for an aco1 gene, and BAC 20 that appeared to have an aco2.
The aco1 3’UTR and aco2 3’UTR probes were used to identify map
positions on the Brassica oleracea integrated map (Sebastion et al., 2000).
There was sufficient segregation of the progeny from the parents in the
doubled haploid lines to calculate a map position for aco1 on linkage group 1
(42.9cM) and for aco2 on linkage group 8 (89.3cM). There were two dominant
bands in Figures 5.12-5.15, suggesting two copies for both aco1 and aco2,
although there was not enough segregation within the progeny to map both
these loci.
The conclusions that can be drawn from these data are that aco1 and
aco2 are individual genes with two copies. This is in accordance with Pogson
et al. (1995a), and examples of small aco gene families including tomato (3
genes), petunia (4 genes) and melon (3 genes) (Barry et al., 1996; Tang et
al., 1993; Lasserre et al., 1996). There is no evidence for an third ACC
oxidase, and aco1 and aco2 may be physically linked in the genome. There
appears to be high sequence conservation between the two Brassica olercea
var. italica cDNA exons, and divergence between the introns. There also
189
appears to be high conservation of the nucleotide sequences of the ACC
oxidases within the Brassica genus, and it is not surprising that the Brassica
napus (AACC) aco gene introns shared 93% nucleotide identity with B.
oleracea (CC).
Genetic and protein conservation suggest an important role for these
genes, such as post-pollination ethylene bursts and enhancing responses to
stress (Rays et al., 2000; Abeles, 1973). This work has shown that there is
nucleotide sequence divergence between the introns of B. oleracea aco1 and
aco2 genes, but not in the exons. There are clearly two copies of the genes,
but are these regulated in the same way, and are they both active? It would
be useful to isolate the promoters and carry out expression studies of these
genes to determine allelic variation. Alleles that produce less post-harvest
ethylene could be selected for to possibly extend broccoli shelf-life.
190
6.0 General Discussion
191
6.1 Assessing the strategies 6.1.1. Constructs
The Brassica oleracea L. var. italica ACC oxidases 1 and 2 and ACC
synthase cDNAs were ligated into the minimal T-DNA vector pSCV1.0
between a CaMV 35S promoter and a nos terminator. It was essential to
produce constructs containing aco1, aco2 and acs in sense and antisense
orientations to understand their role in post-harvest ethylene production. Both
antisense and sense constructs have been shown to down-regulate genes
(Meyer and Saedler, 1996) and sense constructs may even up-regulate the
genes. It would not have been possible to transform broccoli with ACC cDNAs
or to analyse ethylene production in relation to chlorophyll loss without the
constructs. There were a number of factors limiting the cloning strategy, most
notably the large number of restriction sites within the cDNA clones and
cloning vector, using a minimal T-DNA vector and leaving the vector with
restriction sites to change the promoter at some later date. The most efficient
way to achieve this was by amplifying the sequences by PCR and blunt-
ended ligating them into the pJ1.0 backbone vector. Although this approach
was successful for aco1 and aco2, it did not work for ACC synthase. BamHI
sites were added to the primers before PCR and then the product digested
and ligated into the pJ1.0 vector. It was advantageous to have compatible
ends so the cDNA insert could be ligated into the vector in either antisense or
sense orientation. A gus control construct was produced to determine the
impact of the transformation effects on the T0 plants, and confirm activity of
the CaMV 35S promoter. It took longer to produce seven, marker-free
constructs, but these aspects became useful down-stream. In chapter 3,
192
marker-free plants were produced from the progeny of 28/00 aco1A 3 by
selecting out the unwanted rol genes of the Ri TL-DNAs. A marker-free
strategy would be essential to produce commercial transgenic plants, as
these would not be permitted to contain superfluous sequences such as
antibiotic markers. Unfortunately, there was not enough time to grow the T1
generation and assay for the post-harvest production of ethylene and
chlorophyll levels. The ACC synthase sense line in chapter 4 gave information
that would not have been obtained from antisense constructs. It showed that
ACC synthase appears to be the rate-limiting enzyme in the biosynthesis of
ethylene as the buds at 0h produced 245nl/g bud/h ethylene, 5-fold more than
the normal basal level. If the construct was actively producing mRNA
transcripts that were being translated into protein enzymes, it is possible that
more ACC substrate was produced for the ACO to catalyse into ethylene. It
was also important to produce constructs with the native ACC oxidase and
ACC synthase cDNAs isolated by Pogson et al. (1995a,b) from Brassica
oleracea L. var. italica. Henzi et al. (1999b) had obtained the pTOM13
construct (Hamiliton et al., 1990) containing a tomato ACC oxidase between
the CaMV 35S promoter and its terminator. The Brassica oleracea L. var.
italica aco1 and aco2 shared 71% and 70% nucleotide identity with the tomato
ACC oxidase. However, Henzi et al. (1999b) did not achieve sufficient
reduction of ethylene and even increased the ethylene peak at 26h in most of
the lines.
193
6.1.2 Transformation and regeneration
A transformation protocol was chosen to utilise the natural capacity of
Agrobacterium rhizogenes to transfer T-DNA into the recipient plant genome.
The overall transformation efficiency was very low (3.3%), but this could be
improved by inoculating the cut surface soon after excision. Transformation
was highest for the gus construct with 11.7%, suggesting that the ACC
constructs may have had a detrimental effect on the transformation efficiency.
There is no evidence that transformation with another method such as
biolistics, protoplast transformation or Agrobacterium tumefaciens would
produce higher frequencies (Puddephat et al., 1996). Ninety six percent of the
roots transformed with the A. rhizogenes strain LBA 9402 carrying the co-
integrate vector pRi1855::GFP, had been co-transformed with the TL-DNA
and binary T-DNA, containing the gus constructs. The aim to produce 20
independently-transformed GDDH33 root lines for each construct was
achieved. Cells transformed by A. rhizogenes were easily distinguishable by
the emergence of ‘green fluorescing’ hairy roots. Hairy roots are clonal,
arising from single cells (Tempe and Casse-Delbart, 1987) and do not create
chimeric plants. The transformed roots were easily identified and transferred
to media for maintenance and regeneration regeneration. The other
transformation techniques do not transfer the Ri T-DNA containing the rol
genes and therefore have advantages over A. rhizogenes-mediated
transformation. The Ri T-DNA produces a phenotype that exhibits wrinkled
leaves, shortened internode lengths, non-geotropic roots, reduced apical
dominance, altered flower morphology, and reduced seed production (Tepfer,
1990). Severe rol phenotype was present in 4 out 18 of the regenerated T0
194
lines. This did not appear to be related to copy number and relative severity
was probably a ‘position effect’. No seeds were harvested from T0 lines with
severe rol phenotype as the majority of the floral buds had been removed for
ethylene and chlorophyll samples. Seeds were obtained from 11 of the lines
with moderate/weak rol phenotype as these produced larger heads with
greater numbers of floral buds. The results in chapter 5 showed that the
impact of the rol genes reduced both the post-harvest production of ethylene
and chlorophyll loss from broccoli buds. However, it was possible to analyse
these plants when compared to the gus control lines.
In the T1 generation, it will be possible to segregate out the unwanted
rol genes and produce homozygous lines for the construct T-DNA. T1 lines
from different constructs and transformation events could be crossed to
produce lines with stacked genes that may possess a more extreme
phenotype, and produce broccoli heads with a longer post-harvest shelf-life.
Regeneration was clearly the main limiting factor in this process as
from 150 transgenic roots, only 18 regenerated into mature whole plants, and
this took on average a year and a month. It had taken 149 55d from
inoculating to transferring the gus plantlets into the glasshouse, whereas it
had taken 291 21d for the other lines.
The doubled-haploid cultivar GDDH33 was chosen for transformation as it has a homozygous genetic background, is amenable to transformation
through Agrobacterium rhizogenes (H. Robinson, pers. comm. HRI-
Wellesbourne) and is responsive to microspore culture (L. Harvey, Pers.
comm. HRI-Wellesbourne), and also had a short post-harvest head shelf-life.
It was amenable to transformation through A. rhizogenes but produced roots
195
with very poor growth, which often died after a few subcultures. This was
clearly a genotypic effect as earlier experiments with cultivars Shogun and
Packman (results not shown) had produced rapidly growing roots. The
regenerated plants from GDDH33 produced very small heads, limiting
sampling to just ethylene and chlorophyll measurements. It was not possible
to measure the biochemical activity of ACC Oxidase or ACC Synthase in vivo
due to lack of material.
6.1.3 Ethylene and chlorophyll
The strategy for analysing T0 plants was to measure ethylene
production and chlorophyll levels on broccoli buds, the organs that were most
affected by post-harvest senescence. It could be used to test the hypothesis
that endogenous ethylene plays a major a role on sepal chlorophyll loss, and
whether it was possible to extend shelf-life. There was not enough material to
test for the biological activity of ACC Oxidase and ACC Synthase, or to
quantify amounts of proteins, carbohydrates or organic acids. There was not
enough time to analyse mRNA expression levels, as this also would have
been useful.
There was a positive correlation between ethylene production and
chlorophyll loss of post-harvest broccoli buds at 20oC. As ‘total’ ethylene
increased, chlorophyll loss increased. This linear correlation was described by
the equation y = 0.2386x – 23.041, where y = chlorophyll loss and x, the total
production of ethylene. It was confirmed by comparing the ratio of the
regression and the residual mean squares with the F-distribution (F= 0.0015).
196
All of the ACC oxidase constructs (aco1 and aco2, sense and
antisense) significantly (p<0.005) reduced post-harvest production of ethylene
at 24h and significantly (p<0.005) delayed chlorophyll loss, compared to the
untransformed control. The ACC synthase constructs significantly reduced
post-harvest ethylene production at 24h, but did not inhibit a large peak at
72h, and did not significantly reduce chlorophyll loss.
The ‘total’ post-harvest endogenous ethylene production and
chlorophyll loss in the sepals was described by a possible linear relationship.
The relationship also shows that down-regulating the native ACC oxidases
may delay chlorophyll loss. Delaying chlorophyll loss could have a major
impact on retailers as heads producing less ethylene could be stored at 20oC
for longer, and less would go to waste. It could also benefit the consumers, as
the freshness of the head would last longer after purchase.
6.1.4 The aco1 and aco2 genes
In chapter 5, it was shown that there were two copies of aco1 and
aco2, and apparently no other similar aco genes. One of the aco1 loci was
mapped on linkage group 1 and one of the aco2 loci was mapped to linkage
group 8. In both chapters 3 and 5 there were difficulties in achieving
consistent results with the hybridisations due to DNA quantity and purity,
specificity of the probe and hybridisation conditions. In chapter 3, at 65oC with medium stringency washings, the 3’UTR aco1 and aco2 probes did not
hybridise to the native GDDH33 genes. In chapter 5, at 55oC hybridisation
and low stringency washings, the same probes hybridised to the map filters
and the BAC EcoRI RFLP filters. In both sets of experiments there were
197
examples where the probes had not hybridised but there was other evidence
of the correct sequences (PCR, GFP assay, sequencing).
6.2 A role for ethylene
The plant hormone ethylene plays a major role in plant development. In
these experiments it appeared that ethylene increased respiration and plant
growth/senescence. The transformation rate and root growth was
considerably lower in the lines transformed with the ACC constructs
compared to those transformed with the gus constructs. The lines with the gus
constructs significantly regenerated faster (p<0.001) than those with the ACC
constructs. However, the lines with the gus constructs also senesced faster
than the lines with the ACC constructs.
Endogenous ethylene production in the plant appears to be regulated
by a mutligene family of ACC synthases (>1) and two ACC oxidases. A model
was put forward in chapter 4, suggesting that the basal amount of ethylene
produced by the plant is controlled by ACC synthases by producing the ACC
substrate that is oxidised into ethylene by ACO1. It is hypothesised that after
pollination there are two bursts of ethylene, regulated by two ACC synthases
and oxidised by ACO1 and ACO2 in the stamen and carpel structures. This
serves to recycle nutrients for the growing embryo by up-regulating genes
through a phoshorylation pathway that catabolise macromolecules to be
respired for energy. The post-harvest stress on the head, possibly through
lack of photosynthate to the buds appears to up-regulate the aco1 and aco2
and ACC synthase 1 and 2(?), which forces the head into premature
senescence.
198
6.3 Future work 6.3.1 Fully understand the role and expression of the aco and acs genes
Introns 1 and 2 of B. oleracea ACC oxidases 1 and 2 were sequenced
(chapter 5). ACC oxidases 1 and 2 are two distinct genes, each with two loci.
Pogson et al. (1995a) carried out post-harvest aco mRNA expression studies
but little is known of their specific regulation during seed germination,
hypocotyl growth, pollination, stress responses or general maintenance. Only
one ACC synthase has been isolated from Brassica oleracea L var. italica.
ACC synthases are known to belong to multigene families (Rottman et al.,
1991) and are the rate limiting enzymes in the biosynthesis of ethylene (Yang
and Hoffman, 1994). There are at least three different classes of ACC
synthase in tomato, which could be used to probe either cDNA or BAC
libraries of B. oleracea and isolate and sequence the genes (Yang and
Oetiker, 1998). ACC synthase cDNA clones could then be used as probes for
measuring mRNA at different physiological stages or isolating the promoters
and fusing different sections to reporter genes such as gus (Jefferson et al.,
1987) to determine tissue specific expression. 6.3.2 Biochemical aspect
The biochemical activity of ACC synthase and ACC oxidase in pre/post
harvest broccoli and after pollination could be measured. The substrates for
ACC sSynthase, S-adenosylmethionine (SAM) and ACC oxidase
aminocyclopropane carboxylic acid (ACC) could be added to media at
different amounts with broccoli buds and ethylene production measured. The
biochemical activity of the ACC enzymes could be quantified by determining
199
the maximum rate of ethylene production. There could be assays for proteins,
carbohydrates and organic acids (according to King and Morris, 1994), to
confirm the extended quality by retaining nutritious macromolecules.
6.3.3 Improved silencing of aco and acs in broccoli
There is a strong body of evidence that the most effective method for
gene silencing is to introduce tandem inverted DNA repeats into the plant
genome. The mRNA transcripts are thought to hybridise creating double-
stranded RNA molecules (Muskens et al., 2000; Cogoni and Macino, 2000;
Levin et al., 2000). These double-stranded RNAs are thought either to affect
RNA stability, transcription and/or translation directly, or to generate a signal
for gene silencing and defence against viruses (Terryn and Rouze, 2000).
Constructs with inverted repeats of ACC oxidase 1 or 2 and ACC
synthase could be very effective at down-regulating the expression of these
genes. The constructs would have to be targeted at post-harvest expression
or they might have detrimental effects on plant physiology, such as growth
and seed production. A post-harvest specific promoter, preferably one that
has not been patented, would be most suitable. Either the B. oleracea ACC
oxidase 2 or ACC synthase promoters would probably be effective, but the
sequence that up-regulates the genes after pollination would have to be
removed to avoid seed yield loss. Such a construct would not interfere with
any essential requirements, except possibly pathogen attack. A
transformation system with A. rhizogenes could be used to produce marker-
free, phenotypically normal plants.
200
6.3.4 Conventional breeding
In melon (Cucumis melo L. melo) RFLP polymorphisms have been
linked to the length of shelf-life (Zheng and Wolff, 2000). Markers for longer
shelf-lives were associated with low ethylene and markers for shorter shelf-
lives with high ethylene production. It is possible that markers linked to
ethylene production could be used in broccoli to select for lines which produce
lower amounts of ethylene and that would posses longer shelf-lives. This
approach would be more acceptable to the consumers, and possible if there
exists variation between the ACC oxidase and ACC synthase promoters of
different Brassica oleracea L var. italica cultivars.
At HRI-Wellesbourne, there is currently a project to generate a
doubled-haploid mapping population of crosses between the long shelf-life
Mar34 and short shelf-life GD33, for a QTL analysis on bud yellowing and loss
of turgor.
6.3.5 Other ways to improve shelf-life
Application of ethylene biosynthesis inhibitors such as
aminoethoxyvinylglycine (Wang, 1977) and ethylene action including silver
ions (Aharoni et al., 1985) have been shown to delay chlorophyll loss of
broccoli sepals, but are not registered for use on broccoli. Ku and Wills (1999)
have shown that methylcyclopropene, an inhibitor of ethylene action could be
used to achieve such an end, although this would require continuous
exposure to the gas in sealed storage.
Cytokinins have been shown to prolong shelf-life of broccoli by
reducing chlorophyll loss. This was first demonstrated by Batal et al. (1981) by
201
dipping broccoli heads in 6-benzyladenine and supported by Rushing (1990)
and Clarke et al. (1994). Gan and Amasino (1995) isolated a senescence-
associated-gene promoter from Arabidopsis and fused it to an
isopentyltransferase gene (ipt). The gene product from ipt catalyses the rate
limiting step in cytokinin biosynthesis. Gan and Amasino (1995) transformed
tobacco and retarded leaf senescence. Chen et al. (2001) have recently
obtained this construct from Gan and Amasino (1995) and transformed
broccoli with Agrobacterium tumefaciens. They found that bud yellowing was
retarded by >50% over 4 days of post-harvest storage at 25oC. This procedure could be optimised to produce broccoli with greater shelf-life.
This project has shown that the technology exists to successfully
transform crop plants with traits beneficial for the retailer and consumer. It is
essential to understand the genetics and biochemistry behind plant physiology
for the improvement of crop plants, which may be achieved by classical
breeding and molecular techniques.
202
7.0 Appendices
203
7.1 Media
The following media are made up to a litre with R.O. water, and 6g/l
agar added when appropriate.
LB (pH 7.0)
Component (g/l)
yeast extract 10
sodium chloride 10
tryptone 10
YMB (pH 7.0)
Component (g/l)
mannitol 10
yeast extract 0.4
sodium chloride 0.1
magnesium sulphate 0.2
di-potassium hydrogen
orthophosphate
0.5
MGL (pH 7.0)
Component (g/l)
mannitol 5 yeast extract 2.5
bactotryptone 5
sodium chloride 5
sodium-glutamate 1.16
di-potassium hydrogen orthophosphate 0.25
magnesium sulphate 0.1
biotin 0.001
204
MS30 (pH 5.7) Component (g/l)
Murashige and Skoog medium 1
sucrose 30
Regeneration Medium (pH 5.7) Component (g/l)
Murashige and Skoog medium 1
sucrose 30
benzyl adennine 0.005
naphthalene acetic acid 0.005
Rooting and shoot medium (pH 5.7) Component (g/l)
Gamborg’s B5 medium 1
sucrose 20
indole-3-propionic acid 0.0003
Antibiotics (All antibiotics were filter sterilised before use). Antibiotic Solution
cefotaxime 1g Claforan/10mls H2O
kanamycin 0.5g/10mls H2O
ampicillin 1g/10mls H2O
tetracyclin 1g/10mls H2O
chloramphenicol 1g/10mls H2O
rifampicin 0.5g/10mls methanol
205
Histochemical GUS assay (Jefferson et al., 1987)
Assay buffer (X-Gluc)
2mM X-Gluc in 50mM NaHPO4 buffer, pH7.0 Add 0.1% Triton X-100
Stock Solutions
A. Disodium hydrogen orthophosphate (200mM)
B. Sodium dihydrogen orthophosphate (200mM)
Buffer preparation and use
1. Mix 61mls of stock A with 39mls stock B, to make buffer C. 2. Dissolve 10mg X-Gluc in 2.5ml buffer C. 3. Add 200 l for each root sample.
206
7.2 DNA Sequences pJ1.0 T-DNA region (5’-3’)
Right Border
TTACAACGGTATATATCCTGCCA GGAGATCTGATC
CaMV 35S Promoter
AAGCTTGCAT GCCTGCAGGT CCCCAGATTA GCCTTTTCAA TTTCAGAAAG AATGCTAACC
CACAGATGGT TAGAGAGGCT TACGCAGCAG GTCTCATCAA GACGATCTAC CCGAGCAATA
ATCTCCAGGA AATCAAATAC CTTCCCAAGA AGGTTAAAGA TGCAGTCAAA AGATTCAGGA
CTAACTGCAT CAAGAACACA GAGAAAGATA TATTTCTCAA GATCAGAAGT ACTATTCCAG
TATGGACGAT TCAAGGCTTG CTTCACAAAC CAAGGCAAGT AATAGAGATT GGAGTCTCTA
AAAAGGTAGT TCCCACTGAA TCAAAGGCCA TGGAGTCAAA GATTCAAATA GAGGACCTAA
CAGAACTCGC CGTAAAGACT GGCGAACAGT TCATACAGAG TCTCTTACGA CTCAATGACA
AGAAGAAAAT CTTCGTCAAC ATGGTGGAGC ACGACACACT TGTCTACTCC AAAAATATCA
AAGATACAGT CTCAGAAGAC CAAAGGGCAA TTGAGACTTT TCAACAAAGG GTAATATCCG
GAAACCTCCT CGGATTCCAT TGCCCAGCTA TCTGTCACTT TATTGTGAAG ATAGTGGAAA
AGGAAGGTGG CTCCTACAAA TGCCATCATT GCGATAAAGG AAAGGCCATC GTTGAAGATG
CCTCTGCCGA CAGTGGTCCC AAAGATGGAC CCCCACCCAC GAGGAGCATC GTGGAAAAAG
AAGACGTTCC AACCACGTCT TCAAAGCAAG TGGATTGATG TGATATCTCC ACTGACGTAA
GGGATGACGC ACAATCCCAC TATCCTTCGC AAGACCCTTC CTCTATATAA GGAAGTTCAT
TTCATTTGGA GAGAACACGG GGGACT
Linker
CTAGAGGATCCATGTCGACCCGGGAGCT
nos terminator
CGAATTTCCCCGA TCGTTCAAAC ATTTGGCAAT AAAGTTTCTT AAGATTGAAT CCTGTTGCCG
GTCTTGCGAT GATTATCATA TAATTTCTGT TGAATTACGT TAAGCATGTA ATAATTAACA
TGTAATGCAT GACGTTATTT ATGAGATGGG TTTTTATGAT TAGAGTCCCG CAATTATACA
TTTAATACGC GATAGAAAAC AAAATATAGC GCGCAAACTA GGATAAATTA TCGCGCGCGG
TGTCATCTAT GTTACTAGAT CGGGAA
Left Border
TTCGATA TCATTTACAATTGAATATATCCTGCCGCATCG
207
ACC oxidase 1: cDNA sequence (5’- 3’) (Pogson et al., 1995a) 1 atcaaacaac tagctacttc aaccaaatac tttcaaagaa gagagagaga tggagaagaa
61 cattaagttt ccagttgtag acttgtccaa gctcattggt gaagagagag accaaaccat
121 ggctttgatc aacgatgctt gtgagaattg gggcttcttt gagatagtga accatggttt
181 accacatgat ttgatggaca acgtcgagaa gatgacaaag gaacattaca agatatcaat
241 ggaacaaaag ttcaacgaca tgctcaaatc aaaaggtttg gaaaatcttg agagagaagt
301 tgaggatgtt gattgggaaa gcactttcta ccttcgtcat ctccctcagt ccaatctcta
361 cgacattcct gatatgtctg atgaataccg gacggccatg aaagattttg ggaagagatt
421 ggagaatctt gctgaggatt tgttggatct attgtgtgag aatttagggt tagagaaagg
481 gtacttgaag aaagtttttc atggaacaaa aggtccaacc tttgggacta aggtgagcaa
541 ctatccagct tgtcctaagc cagagatgat caaaggtctt agggcccaca ctgatgcagg
601 aggcatcatc ttgttgtttc aagatgacaa ggtcagtggt ctccagcttc ttaaagatgg
661 tgactggatt gatgttcctc cactcaacca ctctattgtc atcaatcttg gtgaccaact
721 tgaggtgata accaacggca ggtacaagag tgtgatgcat cgtgtggtga ctcagaaaga
781 aggaaacaga atgtcaattg catctttcta caacccggga agcgatgctg agatctctcc
841 agcttcatcg cttgcctgta aagaaaccga gtacccaagt tttgtttttg atgactacat
901 gaagctctat gctggggtca agtttcagcc taaggagcca cggttcgagg caatgaagaa
961 tgctaatgca gttacagaat tgaacccaac agcagccgta gagactttct aaaaacaaag
1021tggagtttga gcgaaagcaa agaaacaaaa atgtgttttg tgttgtgtgt ttacgtcaat
1081aagttaaaga ctgatattat tgttgatata attaagatgt ctggtggtta attgttggtc
1141aatggtgtgt ttaaagtgtg gggtgtttat ttatgtttat ggaagatgat aataataata
1201aaaataaata atatgataac tgttctaaga aaaaaaa
ACC oxidase 1: Restriction endonuclease map
208
ACC oxidase 2: cDNA sequence (5’- 3’) (Pogson et al., 1995a) 1 actaacatca tcaacaacat taagttgcca gtttaaagcg attcataagt ttacatagag
61 agggtctaga gtaatggaga aaaacattaa gtttcctgtt gtagacttgt ctaagctaaa
121 tggtgaagag agagaccaaa ccatggcttt ggtcgacgat gcttgtcaaa actggggctt
181 ctttgagctg ctaaaccatg gaataccata tgatcttatg gacaatattg agaggatgac
241 aaaggaacac tacaaaaaat ttatggaaca aaagttcaaa gaaatgcttc gttccaaaaa
301 tttagatact ctggagactg aagttgagaa tgttgattgg gaaagtactt tcttccttca
361 ccatcttcct cagactaatc tctacgacat cccggatatg tcagatgaat accgagcggc
421 catgaaggac tttgggaaga ggctagagaa cctagcggaa gaactgttgg atttgttgtg
481 tgagaatcta gggttggaga aagggtactt gaagaaggtg tttagtggga caaagggtcc
541 aacttttggg acaaaggtta gcaactatcc accatgtcct aaaccggaga tgatcaaagg
601 gcttagggct cacacggatg caggaggcct catattgcta tttcaagacg ataaggtcag
661 tggtcttcag cttctcaaag atggtgtttg ggttgatgtc cctcctctca aacactccat
721 tgtcatcaat cttggtgacc aacttgaggt gataaccaac gggaagtaca agagcataat
781 gcaccgtgtg atgacacaaa aagaaggaaa caggatgtct atagcgtcgt tctacaaccc
841 tggaagtgat gcggagatct ctcccgcacc atctctagtg gagaaagact ctgatgaata
901 tccgagtttt gtgtttgatg actacatgaa gctatatgct ggagtcaagt ttcagcccaa
961 ggagccacgt tttgaggcga tgaagaatgt tgcagcaacc acggatttga acccggtggc
1021cacagtcgag acgttctaaa tggatgtggg attcttctat gaaggcaaga aagatggggt
1081ttgatatgtg tgtggtgggt aataatgtaa aataagtatt gaagttaatg ttatagtaga
1141atcaagaact aagttgtgat cacttatgaa tctagtgtgg ttaagtgtgg gagtgtttat
1201gcaataattg tattggaaat atgaatttta aag
ACC oxidase 2: Restriction endonuclease map
209
ACC synthase: cDNA sequence (5’- 3’) (Pogson et al., 1995b) 1 atctcaacag aacaaaaaca aacccaactt attaaaaccc cttttgaaga aacaaaaatg
61 gtagctttga ctgcagagaa gcaagaccag aacctactgt cgagaatggc cgccggtgac
121 ggacacggcg agaaatcagc ttatttcgat ggctggaaag cttatgaaga aaacccattt
181 cacccaattg atagacccga tggagttatt cagatgggtc tcgctgaaaa tcagctttgt
241 ggagatttga tgcgtaaatg ggttttagaa cacccggaag cttcgatttg cacagctgaa
301 ggtgtgaatc agttcagcga cattgcgatt tttcaggatt atcatggctt gcctgaattc
361 agacaagctg tagcgaagtt tatggagaag acaagaaaca acaaagtgaa gtttgatcct
421 gaccggatcg tcatgagcgg cggcgcaacc ggagcgcacg agacggttgc tttctgttta
481 gccaatcccg gcgacggttt cttggttccg actccttatt atccagggtt tgatagagat
541 ttgagatgga gaaccggagt gaatcttgta ccggttactt gtcatagctc caacgggttt
601 aagatcacgg cggaagcctt ggacgctgcg tacgaaaacg cgcgtgtatc gaacattccg
661 gttaagggtt tactcataac caatccttcc aacccgctcg gtacgacgtt agaccgggat
721 tgcttgaaat ctttggttaa gttcaccaat gacaaaggga ttcaccttat tgctgatgag
781 atctatgcag caactacttt tggtgaatcc gagtttatca gcgtcgcaga agtaattgac
841 gagatccccg attgcaacac cgatttgatc catattgtct atagtttatc aaaagatatg
901 ggtctgcccg gtttaagagt cggtatagta tactcttaca atgaccgggt ggttcaaatt
961 gccaggaaaa tgtcgagttt cggtttggtt tcgtctcaaa cgcagcatct gatcgccaaa
1021atgttatccg acgaagactt tgtagacgaa ttcatccgca agagcaaact acggttggct
1081gaaagacacg cagagttaac caccggttta gacggtttaa gcattggttg gttaaaggcc
1141ggagccggtt tgttcatatg gatggattta agaaaccttt tgaagacggc tacattcgac
1201tcggagatgg agctgtggcg tgtgatcgtc cacaaggtga agctgaacgt ttctccaggc
1261ggttcgtgcc actgccacga accgggatgg tttagagtat gttttgcgaa catggactac
1321cagacaatgg agaccgctct agagaggatc agagtgttca ctagtcaaac tgaagaggag
1381agtctgatac tgactaaacc gatggctaag aaaaagaagt gttggcagag tagcctccgg
1441ttaagcttta aggacacgag acggttcgag gagggcttct tctctcctca ttcaccggtg
1501ccaccttctc cgcttgttcg tgcacagata taagaccgtt tcagtttttg actagaccaa
1561gctgtcgtaa attgaaagat aaaaattatt taacattgat tgtgacattt tgtctaatta
1621aatgtatagc tactactgtc aagttgaata tttttataat gtatctcttg taagtttaac
1681cagttgaata atataatatg agaaaccata atttctctaa tcgttattaa aaaa
ACC synthase: Restriction endonuclease map
210
8.0 References
211
Abeles, F.B. (1973) Ethylene in Plant biology. New York: Acacdemic. pp. 302 Aharoni, N., Philosoph-Hadas, S., Barkai-Golan, R. (1985) Modified atmospheres to delay senescence and decay of Broccoli. In SM Blankenship, ed, Controlled Atmospheres for Storage and Transport of Perishable Agricultural Commodities, Horticulture Report 126. Department of Horticultural Science, Raleigh, NC, pp 167-177
Al-Kaff, N.S., Covey, S.N., Kreike, M.M., Page, A.M., Pinder, R., Dale, P.J. (1998) Transcriptional and posttranscriptional plant gene silencing in response to a pathogen. Science, 279: 2113-2115
Aronson, A.I., Shai, Y. (2001) Why Bacillus thuringiensis insecticidal toxins are so effective: unique features of their mode of action. FEMS Microbiology Letters, 195: 1-8
Ayub, R., Guis, M., BenAmor, M., Gillot, L., Roustan, J.P., Latche, A., Bouzayen, M., Pech, J.C. (1996) Expression of ACC oxidase antisense gene inhibits ripening of cantaloupe melon fruits. Nature Biotechnology, 14: 862-866
Barry, C.S., Blume, B., Bouzayen, M., Cooper, W., Hamilton, A.J., Grierson, D. (1996) Differential expression of the 1-aminocyclopropane-1-carboxylate oxidase gene family of tomato. Plant Journal, 9: 525-535
Batal, K.M., Heaton, E.K., Beuchat, L.R., Granberry, D.M. (1981) Effects of N- 6-benzyladenine and storage temperatures on shelf-life and quality of raw and cooked broccoli (Brassica oleracea L var italica). Hortscience, 16: 284-284
Benfy, P.N., Chua, N.-H. (1990) The cauliflower mosaic virus 35S promoter: Combinational regulation of transcription in plants. Science, 250: 959-966
Berthomieu, P., Beclin, C., Charlot, F., Dore C., Jouanin L. (1994) Routine transformation of rapid cycling cabbage (Brassica oleracea)-molecular evidence for regeneration of chimeras. Plant Science, 96: 223-235
Bevan, M. (1984) Binary Agrobacterium vectors for plant transformation. Nucleic Acids Research, 12: 8711-8721
Bevan, M., Barnes, W.M., Chilton, M-D. (1983) Structure and transcription of the nopaline synthase gene region of the T-DNA. Nucleic Acids Research, 11: 369-385
Blaich, R. (1992) Function of genetic material: activity of genes in transgenic plants. Progress in Botany, 53: 207-220
Blumenthal, A., Kuznetzova, L., Edelbaum, O., Raskin, V., Levy, M., Sela, I. (1999) Measurement of green fluorescence protein in plants: quantification, correlation to expression, rapid screening and differential gene expression. Plant Science, 142: 93-99
212
Bolitho, K.M., LayYee, M., Knighton, M.L., Ross, G.S. (1997) Antisense apple ACC-oxidase RNA reduces ethylene production in transgenic tomato fruit. Plant Science, 122: 91-99
Braun, A. (1954) The physiology of plant tumors. Annual Review of Plant Physiology, 5: 133-162
Braun, A. (1978) Plant tumours. Biochimica Et Biophysica Acta, 51: 167-191
Brennan, P.S., Shewfelt, R.L. (1989) Effect of cooling delay at harvest on broccoli quality during postharvest storage. Journal of Food Quality, 12: 13-22
Bryant, J.A. (1988) Plants Today, 1: 23-28
Buchanan-Wollaston, V. (1997) The molecular biology of leaf senescence. Journal of Experimental Botany, 48: 181-199
Burg, S.P. (1962) The physiology of ethylene formation. Annual Review of Plant Physiology, 13: 265-302
Buttner, M., Singh, K.B. (1997) Arabidopsis thaliana ethylene-responsive element binding protein (AtEBP), an ethylene inducible, GCC box DNA- binding protein interacts with an ocs element binding protein. Proceedings of National Academy of Science USA, 94: 5961-5966
Cardarelli, M., Mariotti, D., Pomponi, M., Spano, L., Capone, I., Constantino, P (1987) Agrobacterium rhizogenes T-DNA genes capable of inducing hairy root phenotype. Molecular and General Genetics, 209: 475-480
Chalfie, M., Tu, Y., Euskirchen, G., Ward, W.W., Prasher, D.C. (1994) Green fluorescent protein as a marker gene for gene expression. Science, 263: 802- 805
Charest, P.J., Holbrook, L.A., Gabard, J., Iyer, V.N., Miki, B.L. (1988) Agrobacterium-mediated transformation of thin cell layer explants from Brassica napus L. Theoretical and Applied Genetics, 75: 438-445
Chen, L.F.O., Hwang, J.Y., Charng, Y.Y., Sun, C.W., Yang, S.F. (2001) Transformation of broccoli (Brassica oleracea var. italica) with isopentenyltransferase gene via Agrobacterium tumefaciens for post-harvest yellowing retardation Molecular Breeding, 7: 243-257
Chi, G.L., Barfield, D.G., Sim, G.E., Pua, E.C. (1990) Effect of AgNO3 and amino-ethoxyvinylglycine on in vitro shoot and root organogenesis from seedling explants of recalcitrant brassica genotypes Plant Cell Reports, 9: 195-198
Chrispeels, M.J., Sadava, D.E. (1994) Plants, Genes and Agriculture. Jones & Bartlett Pub.
213
Christey, M.C., Braun, R.H., Reader, J.K. (1999) Field performance of transgenic brassicas (Brassica oleracea and B. rapa) transformed with Agrobacterium rhizogenes. SABRAO Journal of Breeding and Genetics, 31:93-108
Christey, M.C., Sinclair, B.K. (1992) Regeneration of transgenic kale (Brassica oleracea var acephala), rape (Brassica napus) and turnip (Brassica campestris var rapifera) plants via Agrobacterium rhizogenes mediated transformation. Plant Science, 87: 161-169
Christou, P. (1996) Transformation technology. Trends In Plant Science, 1: 423-431
Clarke, S.F., Jameson, P.E., Downs, C.G. (1994). The influence of 6- benzylaminopurine on post-harvest senescence of floral tissues of broccoli (Brassica oleracea var. italica). Plant Growth Regulation, 14: 21-27
Cogan, N.O.I., Harvey, E., Robinson, H.T., Lynn, J., Pink, D.A.C., Newbury, H.J., Puddephat, I.J. (2001) The effects of anther culture and plant genetic background on Agrobacterium rhizogenes-mediated transformation of commercial cultivars and derived double haploid Brassica oleracea. Plant Cell Reports, 20:755-762
Cogoni, C., Macino, G. (2000) Post-transcriptional gene silencing across kingdoms Current Opinion in Genetics & Development, 10: 638-643
Cordes, S., Deikmann, J., Margossian, L.J., Fischer, R.L. (1989) Interaction of a DNA binding factor with the 5 flanking region of an ethylene-responsive fruit ripening gene from tomato. EMBO Journal, 7: 3315-3320
Crisp, P., Gray, A.R., Angell, S.M., Salter, P.J., Akehurst, J., Sutherland, R.A (1986). The effect of spacing on the breeding of broccoli from an expanded genetic base. Journal of Horticultural Science, 61: 205
Daley, M., Knauf, V.C., Summerfelt, K.R., Turner, J.C (1998) Co- transformation with one Agrobacterium tumefaciens strain containing two binary plasmids as a method for producing marker-free transgenic plants. Plant Cell Reports, 17: 489-496
David C., Tempe, J. (1988) Genetic transformation of cauliflower (Brassica oleracea L. var. botrytis) by Agrobacterium rhizogenes. Plant Cell Reports, 7: 88-91
DEFRA (2000) Basic Hroticultural Statistics for the United Kingdom, Calender and Crop Years 1989/90-1999/00
de Framond, A.J., Back, E.W., Chilton, W.S., Kayes, L., Chilton, M-D. (1986) Two unlinked T-DNAs can transform the same tobacco plant cell and segregate in the F1 generation. Molecular and General Genetics, 202: 125- 131
214
Delagrave, S., Hawtin, R.E., Silva, C.M., Yang, M.M., Youvan, D.C. (1995) Red-shifted excitation mutants of the green fluorescent protein. Bio/Technology, 13: 151-154
Depicker, A., Herman, L., Jacobs, A., Schell, J., Van Montagu, M. (1985) Frequencies of simultaneous transformation with different T-DNAs and their relevance to the Agrobacterium/plant cell interaction. Molecular and General Genetics, 201: 477-484
Draper, J., Scott, R., Armitage, P., Walden, R. (1988) Plant genetic transformation and gene expression: A laboratory manual. Blackwell Scientific pub. Oxford.
Earle, E.D., Metz T.D., Roush R.T., Shelton A.M. (1996) Advances in transformation technology for vegetable brassica. Acta Horticulturae, 407: 161-168
Felix, G., Grosskopf , D.G., Regenass, M., Basse, C.W., Boller, T. (1991) Elicitor-induced ethylene biosynthesis in tomato cells - characterization and use as a bioassay for elicitor action. Plant Physiology, 97: 19-25
Filichkin, S.A., Gelvin, S.B. (1993) Formation of a putative relaxation intermediate during T-DNA processing directed by the Agrobacterium tumefaciens VirD1, D2 endonuclease. Molecular Microbiology, 8: 915-926
Fray, R.G., Grierson, D. (1993). Molecular genetics of tomato fruit ripening. Trends in Genetics, 9: 438-443
Frisch, D.A., Harris-Haller, L.W., Yokubaitis, N.T., Thomas, T.L., Hardin, S.H., Hall, T.C. (1995). Complete sequence of the binary vector Bin 19. Plant Molecular Biology, 27:405-409
Fry, J., Barnason, A., Horsch, R.B (1987) Transformation of Brassica napus with Agrobacterium tumefaciens based vectors. Plant Cell Reports, 6: 321-325
Gan, S.S., Amasino, R.M. (1995) Inhibition of leaf senescence by autoregulated production of cytokinin Science, 270 (5244): 1986-1988
Gelvin, S.B. (1990). Crown Gall Disease and Hairy Root Disease. Plant Physiology, 92: 281-285
Gelvin, S. B. (2000). Agrobacterium and plant genes involved in T-DNA transfer and integration. Annual Review of Plant Physiology, 51: 223-256
Gonzalez, N.M and Botella, J.R (2001) Characterization of three ACC synthase gene family members during post-harvest-induced senescence in broccoli (Unpublished).
Gray, A.R. (1982). Taxonomy and evolution of broccoli (Brassica oleracea var
215
italica). Economic Botany, 36: 397 Gray, A.R. (1993). Broccoli Brassica oleraea L. (italica group). In: Genetic improvement of vegetable crops. Eds. Kalloo, G., Bergh, B.O. pp 61-86.
Gray, J., Picton, S., Shabbeer, J., Schuch, W., Grierson, D. (1992) Molecular- biology of fruit ripening and its manipulation with antisense genes. Plant Molecular Biology, 19: 69-87
Grierson, D. (1996). Silent genes and everlasting fruits and vegetables? Nature Biotechnology, 14: 828-829
Guilley, H., Dudley, R.K., Jonard, G., Balazs, E., Richards, K.E. (1982). Transcription of cauliflower mosaic-virus DNA - Detection of promoter sequences, and characterization of transcripts. Cell, 30: 763-773
Guis, M., Botondi, R., BenAmor, M., Ayub, R., Bouzayen, M., Pech, J.C., Latche, A. (1997) Ripening-associated biochemical traits of Cantaloupe Charentais melons expressing an antisense ACC oxidase transgene. Journal of American Society of Horticultural Science, 122: 748-751
Hadfield, K.A., Bennett, A.B. (1997) Programmed senescence of plant organs. Cell Death and Differentiation, 4: 662-670
Hamilton, A.J., Brown, S., Han, Y.H., Ishizuka, M., Lowe, A., Solis, A.G.A., Grierson, D. (1998) A transgene with repeated DNA causes high frequency, post-transcriptional suppression of ACC-oxidase gene expression in tomato. Plant Journal, 15: 737-746
Hamilton, A.J., Lycett, G.W., Grierson, D. (1990). Antisense gene that inhibits synthesis of the hormone ethylene in transgenic plants. Nature, 346: 284-287
Hao, D.Y., OhmeTakagi, M., Sarai, A. (1998) Unique mode of GCC box recognition by the DNA-binding domain of ethylene-responsive element- binding factor (ERF domain) in plants. Journal of Biological Chemistry, 273: 26857-26861
Haseloff J, Siemering KR, Prasher DC, Hodge S (1997) Removal of a cryptic intron and subcellular localization of green fluorescent protein are required to mark transgenic Arabidopsis plants brightly. Proceedings of the National Academy of Sciences USA, 94: 2122-2127
Heim, R., Cubitt, A.B., Tsien, R.Y. (1995) Improved green fluorescence. Nature, 373: 663-664
Heim, R., Prasher, D.C., Tsien, R.Y. (1994) Wavelength mutations and posttranslational autoxidation of green fluorescent protein. Proceedings of the National Academy of Sciences USA, 91: 12501-12504
Helm, J. (1963) Morphologisch-taxonomische Gleiderung der Kultursippen
216
von Brassica oleracea. Kulturpflanze, 11: 92-210 Henzi, M.X., Christey, M.C., McNeil, D.L. (2000) Factors that influence Agrobacterium rhizogenes-mediated transformation of broccoli (Brassica oleracea L. var. italica). Plant Cell Reports, 19: 994-999
Henzi, M.X., Christey, M.C., McNeil, D.L., Davies, K.M. (1999a) Agrobacterium rhizogenes-mediated transformation of broccoli (Brassica oleracea L var. italica) with an antisense 1-aminocyclopropane-1-carboxylic acid oxidase gene. Plant Science, 143: 55-62
Henzi, M.X., McNeil, D.L., Christey, M.C., Lill, R.E. (1999b) A tomato antisense 1-aminocyclopropane-1-carboxylic acid oxidase gene causes reduced ethylene production in transgenic broccoli. Australian Journal of Plant Physiology, 26: 179-183
Hipkins, M.F., Baker, N.R. (1986) Spectoscopy. In Photosynthesis energy transduction Ed Hipkins, M.F., Baker, N.R. IRL press Oxford
Hodgkin, T. (1995). Cabbages, Kales, etc. Brassica oleracea (Cruciferae). In: Smartt J. Simmonds NW (eds) Evolution of Crop Plants. pp 76-82. Longman
Hoekema, A., Hirsch, P.R., Hooykaass, P.J.J., Schilperoot, R.A. (1983). A binary vector strategy based on separation of vir- and T-region of Agrobacterium tumefaciens Ti-plasmid. Nature, 303: 179-180
Hooykaas, P.J.J., Beijersbergen, A.G.M (1994) The virulence system of Agrobacterium tumefaciens. Annual Review of Phytopathology, 32: 157-179
Hosoki, T., Kigo, T. (1994) Transformation of brussels sprouts (Brassica oleracea var. gemmifera Zenk.) by Agrobacterium rhizogenes harbouring a reporter -glucuronidase gene. Journal of Japanese Society of Horticultural Science, 63: 589-592
Hosoki, T., Kigo, T, Shiraishi, K. (1991) Transformation and regeneration of broccoli (Brassica oleracea var. italica) mediated by Agrobacterium rhizogenes. Journal of Japanese Society of Horticultural Science, 60: 71-75
Hyodo, H., Morozumi, S., Kato, C., Tanaka, K., Terai, H. (1994) Ethylene production and ACC oxidase activity in broccoli flower buds and the effect of endogenous ethylene on their senescence. Acta Horticulturae, 394: 191-198
Jackson, S.A., Fenggao, D., Jiang, J. (1999) Digital mapping of bacterial artificial chromosomes by fluorescence in situ hybridisation. The Plant Journal, 17: 581-587
James, C. Global review of commercialized transgenic crops: 2000. ISAAA Briefs No. 21
Jefferson, R.A., Burgess, S.M., Hirsh, D. (1986) Beta-glucuronidase from
217
Escherichia coli as a gene-fusion marker. Proceedings of the National Academy of Sciencess USA, 83: 8447-8451
- glucuronidase as a sensitive and versatile gene fusion marker in higher plants. The EMBO Journal, 6:3901-3907
Jin, S., Prusti, R.K., Roitsch, T., Ankenbauer, R.G., Nester, E.W. (1990a) Phosphorylation of the VirG protein of Agrobacterium tumefaciens by the autophosphorylated VirA protein: essential role in biological activity of VirG. Journal of Bacteriology, 172: 4945-4950
Jin, S., Roitsch, T., Ankenbauer, R.G., Gordon, M.P., Nester, E.W. (1990b) The VirA protein of Agrobacterium tumefaciens is autophosphorylated and is essential for vir gene regulation. Journal of Bacteriology, 172: 525-530
John, I., Drake, R., Farrell, A., Cooper, W., Lee, P., Horton, P., Grierson, D. (1995) Delayed leaf senescence in ethylene-deficient ACC-oxidase antisense tomato plants: molecular and physiological analysis. The Plant Journal, 7: 483-490
Jones, M.L., Woodson, W.R. (1999) Differential expression of three members of the 1-aminocyclopropane-1-carboxylate synthase gene family in carnation. Plant Physiology, 119: 755-764
Kasai, Y., Hyodo, H., Ikoma, Y., Yano, M. (1998) Characterization of 1- aminocyclopropane-1-carboxylate (ACC) oxidase in broccoli florets and from Escherichia coli cells transformed with cDNA of broccoli ACC oxidase. Botanical Bulletin Of Academia Sinica, 39: 225-230
Kasai, Y., Kato, M., Hyodo, H. (1996). Ethylene biosynthesis and its involvement in senescence of broccoli florets. Journal of Japanese Society of Horticultural Science, 65: 185-191
Kato, M., Hyodo, H. (1999) Purification and characterization of ACC oxidase and increase in its activity during ripening of pear fruit. Journal of Japanese Society of Horticultural Science, 68: 551-557
Kende, H. (1989) Enzymes of ethylene biosynthesis. Plant Physiology, 91: 1-4
Kende, H. (1993) Ethylene Biosynthesis. Annual Review of Plant Physiology and Plant Molecular Biology, 44: 283-307
Kim, W.T., Yang, S.F. (1992) Turnover of 1-aminocyclopropane-1-carboxylic acid synthase protein in wounded tomato fruit tissue. Plant Physiology, 100: 1126-1131
King, G.A., Morris, S.C. (1994a). Early compositional changes during postharvest senescence of broccoli. Journal of American Society of Horticultural Science, 119: 1000-1005
218
King, G.A., Morris, S.C. (1994b). Physiological-changes of broccoli during early postharvest senescence and through the preharvest-postharvest continum. Journal of American Society of Horticultural Science, 119: 270-275
Knoester, M., Linthorst, H.J.M., Bol, J.F., Van Loon, L.C. (1997) Modulation of stress-inducible ethylene biosynthesis by sense and antisense gene expression in tobacco. Plant Science, 126:173-183
Komari, T., Hiei, Y., Saito, Y., Murai, N., Kumashiro, T. (1996) Vectors carrying two separate T-DNAs for co-transformation of higher plants mediated by Agrobacterium tumefaciens and segregation of transformants free from selection markers. The Plant Journal, 10: 165-174
Konings, H., Jackson, M.B. (1979) A relationship between rates of ethylene production by roots and the promoting or inhibiting effects of exogenous ethylene and water on root elongation. Zeitschrift Fur Pflanzenphysiologie, Bd. 92. S. 385-397
Ku, V.V.V., Wills, R.B.H. (1999) Effect of 1-methylcyclopropene on the storage life of broccoli. Postharvest Biology and Technology, 17: 127-132
Kumar, P.P., Lakshmanan, P., Thorpe, T. (1998) Review Regulation of morphogenesis in plant tissue culture by ethylene. In vitro Cellular and Developmental Biology-Plant, 34: 94-103
Lasserre, E., Bouquin, T., Hernandez, J.A., Bull, J., Pech, J.C., Balague, C. (1996) Structure and expression of three genes encoding ACC oxidase homologs from melon (Cucumis melo L.). Molecular and General Genetics, 251: 81-90
Lawton, K.A., Huang, B., Goldsbrough, P.B., Woodson, W.R. (1989) Molecular-cloning and characterization of senescence-related genes from carnation flower petals. Plant Physiology, 90: 690-696
Lawton, K.A., Raghothama, K.G., Goldsbrough, P.B., Woodson, W.R. (1990) Regulation of senescence-related gene-expression in carnation flower petals by ethylene. Plant Physiology, 93: 1370-1375
Levin, J.Z., de Framond, A.J., Tuttle, A., Bauer, M.W., Heifetz, P.B. (2000) Methods of double-stranded RNA-mediated gene inactivation in Arabidopsis and their use to define an essential gene in methionine biosynthesis. Plant Molecular Biology, 44: 759-775
Liang, X.W., Abel, S., Keller, J.A., Shen, N.F., Theologis, A. (1992) The 1- aminocyclopropane-1-carboxylate synthase gene family of Arabidopsis thaliana. Proceedings of the National Acadamy of Science USA, 89: 11046- 11050
Lieberman, M., Hardenburg, R.E. (1954) Effect of modified atmospheres on
219
respiration and yellowing of broccoli at 75 degrees F. Proceedings of the American Society of Horticultural Science, 63: 409-414
LopezGomez, R., Campbell, A., Dong, J.G., Yang, S.F., GomezLim, M.A. (1997) Ethylene biosynthesis in banana fruit: Isolation of a genomic clone to ACC oxidase and expression studies. Plant Science, 123: 123-131
Makhlouf, J., Willemot, C., Arul, J., Castaigne, F., Emond, J.P. (1989) Regulation of ethylene biosynthesis in broccoli flower buds in controlled atmospheres. Journal of the American Society of Horticultural Science, 114: 955-958
Martintanguy, J., Corbineau, F., Burtin, D., Benhayyim, G., Tepfer, D. (1993) Genetic-transformation with a derivative of rolC from Agrobacterium- rhizogenes and treatment with alpha-aminoisobutyric-acid produce similar phenotypes and reduce ethylene production and the accumulation of water- insoluble polyamine-hydroxycinnamic acid conjugates in tobacco flowers. Plant Science, 93: 63-76
Mattanovich, D., Ruker, F., Machadom A.D., Laimer, M., Regner, F., Steinkellner, H., Himmler, G., Katinger, H. (1989) Efficient transformation of Agrobacterium spp by electroporation. Nucleic Acids Research, 17: 6747- 6747
MbeguieAMbeguie, D., Chahine, H., Gomez, R.M., Gouble, B., Reich, M., Audergon, J.M., Souty, M., Albagnac, G., FilsLycaon, B. (1999) Molecular cloning and expression of a cDNA encoding 1- aminocyclopropane-1- carboxylate (ACC) oxidase from apricot fruit (Prunus armeniaca). Physiologia Plantarum, 105: 294-303
McCormac, A.C., Fowler, M.R., Chen, D-F., Elliot, M.C. (2001) Efficient co- transformation of Nicotiana tabacum by two independent T-DNAs, the effect of T-DNA size and implications for genetic separation. Transgenic Research, 10: 143-155
McGarvey, D.J., Sirevag, R., Christoffersen, R.E. (1992) Ripening-related gene from avocado fruit - ethylene-inducible expression of the messenger- RNA and polypeptide. Plant Physiology, 98: 554-559
McGrath, R.B., Ecker, J.R. (1998). Ethylene signalling in Arabidopsis: Events from the membrane to nucleus. Plant Physiology and Biochemistry, 36: 103- 113
McKeon, T.A., Fernandez-Maculet, J.C., Yang, S.F. (1995) Biosynthesis and metabolism of ethylene. pp. 118-139. In: P.J. Davies (ed). Plant hormones. Kluwer Academic publishers, Dortrecht
Meyer, P., Saedler, H. (1996) Homology-dependent gene silencing in plants. Annual Review of Plant Physiology And Plant Molecular Biology, 47: 23-48
220
Molinier, J., Himber, C., Hahne, G. (2000) Use of green fluorescent protein for the detection of transfromed shoots and homozygous offspring. Plant Cell Reports, 19:219-223
Morgan, P.W., Drew, M.C. (1997) Ethylene and plant responses to stress. Physiologia Plantarum, 100: 620-630
Murashige, T., Skoog F. (1962) A revised medium for rapid growth and bioassays with tobacco tissue culture. Physiologia Plantarum, 15: 473-497
Muskens, M.W.M, Vissers, A.P.A., Mol, J.N.M., Kooter, J.M. (2000) Role of inverted DNA repeats in transcriptional and post-transcriptional gene silencing. Plant Molecular Biology, 43: 243-260
Nichols, R. (1977) Sites of ethylene production in the pollinated and unpollinated senescing carnation (Dianthus caryophyllus) inflorescence. Planta, 135: 155-159
Oeller, P.W., Wong, L.M., Taylor, L.P., Pike, D.A., Theologis, A. (1991). Reversible inhibition of tomato fruit senescence by antisense RNA. Science, 254: 437-439
Oetiker, J.H., Olson, D.C., Shiu, O.Y., Yang, S.F. (1997) Differential induction of seven 1-aminocyclopropane-1-carboxylate synthase genes by elicitor in suspension cultures of tomato (Lycopoersicon esculentum). Plant Molecular Biology, 34: 275-286
Olson, D.C., White, J.A., Edelman, L., Harkins, R.N., Kende, H. (1991) Differential expression of two genes from 1-aminocyclopropane-1-carboxylate synthase in tomato fruits. Proceedings of the National Academy of Sciences USA, 88: 5340-5344
O’Neil, S.D., Nadeau, J.A., Zhang, X.S., Bui, A.Q., Halevy, A.H. (1993) Interorgan regulation of ethylene biosynthetic genes by pollination. The Plant Cell, 5: 419-432
Ow, D.W., Wood, K.V., DeLuca, M., Wet J.R de., Helinski, D.R., Howell, S.H. (1986) Transient and stable expression of the firefly luciferase gene in plant cell and transgenic plants. Science, 234: 856-859
Paterson, A.H., Bowers, J.E., Burrow, M.D., Draye, X., Elsik, C.G., Jiang, C.X., Katsar, C.S., Lan, T.H., Lin, Y.R., Ming, R., Wright, R.J. (2000) Comparative genomics of plant chromosomes. Plant Cell, 12: 1523-1539
Payton, S., Fray, R.G., Brown, S., Grierson, D. (1996) Ethylene receptor expression is regulated during fruit ripening, flower senescence and abscission. Plant Molecular Biology, 31: 1227-1231
Picton, S., Gray, J.E., Grierson, D. (1995) The manipulation of tomato fruit ripening by expression of antisense RNA in transgenic plants. Euphytica, 85:
221
193-202 Pierpoint, W.S. (1995) Targets for the introduction of pest and disease resistance crops by genetic engineering. IACR- Long Ashton Research Station. April 1995
Pogson, B.J., Downs, C.G., Davies, K.M. (1995a). Differential expression of two 1-aminocyclopropane-1-carboxylic acid oxidase genes in broccoli after harvest. Plant Physiology, 108: 651-657
Pogson, B.J., Downs, C.G., Davies, K.M., Morris, S.C. (1995b). Nucleotide sequence of a cDNA clone encoding 1-aminocyclopropane-1-carboxylic acid synthase from broccoli. Plant Physiology, 108: 857-858
Pogson, B.J., Morris, S.C. (1997). Consequences of cool storage of broccoli on physiological and biochemical changes and subsequent senescence at 20oC. Journal of the American Society of Horticultural Science, 122: 553-558
Poulsen, G.B. (1996). Genetic transformation of Brassica. Plant Breeding, 115: 209-255
Prasher, D.C. (1995) Using GFP to see the light. Trends in Genetics, 11: 320- 323
Pua, E.C., Chi, G.L. (1993) De Novo shoot morphogenesis and plant growth of mustard (Brassica juncea) in vitro in relation to ethylene. Physiologia Plantarum, 88: 467-474
Pua, E.C., Sim, G.E., Chye, M.L. (1992) Isolation and sequence analysis of a cDNA clone encoding ethylene forming enzyme in Brassica juncea (L.) Czern & Coss. Plant Molecular Biology, 19: 541-544
Puddephat, I.J., Riggs, T.J., Fenning, T.M. (1996). Transformation of Brassica oleracea L.: A critical review. Molecular Breeding, 2: 185-210
Puddephat, I.J., Robinson, H.T., Fenning, T.M., Barbara, D.J., Morton, A., Pink, D.A.C. (2001) Recovery of phenotypically normal transgenic plants of Brassica oleracea upon Agrobacterium rhizogenes-mediated co- transformation and selection of transformed hairy roots by GUS assay. Molecular Breeding, 7: 229-242
Purnhauser, L., Medgyesy, M., Czeko, P.J., Marton, L. (1987) Stimulation of shoot regeneration in Triticum aestivum and Nicotiana plumbaginifolia Viv. tissue culture using the ethylene inhibitor AgNO3. Plant Cell Reports, 6: 1-4
Rays, D.B., Reid, D.M., Yeung, E.C., Pharis, R.P. (2000) Role of ethylene in cotyledon development of microspore-derived embryos of Brassica napus. Journal of Experimental Botany, 51: 1851-1859
Robinson, K.E.P., Adams, D.O. (1987) The role of ethylene in the
222
regeneration of Helianthus annuus (sunflower) plants from callus. Physiologia Plantarum, 71: 151-156
Rottmann, W.H., Peter, G.F., Oeller, P.W., Keller, J.A., Shen, N.F., Nagy, B.P., Taylor, L.P., Campbell, A.D., Theologis, A. (1991) 1-aminocyclopropane- 1-carboxylate synthase in tomato is encoded by a multigene family whose transcription is induced during fruit and floral senescence. Journal of Molecular Biology, 222: 937-961
Rushing, J.W. (1990) Cytokinins affect respiration, ethylene production, and chlorophyll retention of packaged broccoli florets. Horticultural Science, 25: 88-90
Sambrook, J., Fritsch, E.F., Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual. 2nd. Edition. Cold Spring Harbor Laboratory Press, Plainview, New York.
Sato, T., Theologis, A. (1989). Cloning the mRNA encoding 1- aminocyclopropane-1-carboxylate synthase, the key enzyme for ethylene biosynthesis in plants. Proceedings of the National Academy of Sciences USA, 86: 6621-6625
SatoNara, K., Yuhashi, K., Higashi, K., Hosoya, K., Kubota, M., Ezura, H. (1999) Stage- and tissue-specific expression of ethylene receptor homolog genes during fruit development in muskmelon. Plant Physiology, 120: 321-329
Schmulling, T., Schell, J., Spena, A. (1988) Single genes from Agrobacterium rhizogenes influence plant development. The EMBO Journal, 7: 2621-2629
Sebastion, R.L., Howell, E.C., King, G.J., Marshall, D.F., Kearsey, M.J. (2000) An integrated AFLP and RFLP Brassica oleracea linkage map from two morphologically distinct doubled-haploid mapping populations. Theoretical Applied Genetics, 100: 75-81
Senior, I., Holford, P., Cooley, R.N., Newbury, H.J. (1995) Transformation of Antirrhinum majus using Agrobacterium rhizogenes. Journal of Experimental Botany, 46: 1233-1239
Sethi, U., Basu, A., Guha-Mukherjee, S. (1990) Control of cell proliferation and differentiation by modulators of ethylene biosynthesis and action in Brassica hypocotyl explants. Plant Science, 69: 225-229
Sheen, J., Hwang, S., Niwa, Y., Kobayashi, H., Galbraith, D.W. (1995) Green- fluorescent protein as a new vital marker in plant cells. The Plant Journal, 8: 777-784
Sinkar, V.P., Pythoud, F., White, F.F., Nester, E.W., Gordon, M.P. (1988) rolA locus of the Ri plasmid directs developmental abnormalities in transgenic tobacco plants. Genes and Development, 2: 688-697
223
Slater, A., Maunders, M.J., Edwards, K., Schuch, W., Grierson, D. (1985) Isolation and characterisation of cDNA clones for tomato polygalcturonase and other ripening-related proteins. Plant Molecular Biology, 5: 37-147
Slightom, J.L., Durand-Tardif, M., Jouanin, L., Tepfer, D. (1986). Nucleotide sequence analysis of TL-DNA of Agrobacterium rhizogenes agropine type plasmid. Journal of Biological Chemistry, 261: 108-121
Song, K., Osborn, T.C., Williams, P.H. (1988) Brassica taxonomy based on nuclear restriction fragment length polymorphisms (RFLPs), 3. Genome relationships in Brassica and related genera and the origin of B. oleracea and B. rapa (syn. Campestris). Theoretical and Applied Genetics, 79: 497
Songstad, D.D., Duncan, D.R., Widholm, J.M. (1988) Effect of 1- aminocyclopropane-1-carboxylic acid, silver nitrate, and norbornadiene on plant regeneration from maize callus cultures. Plant Cell Reports, 7: 262-265
Spano, L., Pomponi, M., Constantino, P., Van Slogteren, G.M.S., Tempe, J. (1982) Identification of T-DNA in the root-inducing plasmid of the agropine type Agrobacterium rhizogenes 1855. Plant Molecular biology, 1: 291-300
Spanu, P., Grosskopf, D.G., Felix, G., Boller, T. (1994) The apparent turnover of 1-aminocyclopropane-1-carboxylate synthase in tomato cells is regulated by protein-phosphorylation and dephosphorylation. Plant Physiology, 106: 529-535
Stam, P., Van Ooijen, J.W. (1995) JoinMap (Tm) 2.0: software for the calculation of genetic maps. CPRO-Wageningen
Tang, X., Wang, H., Brandt, A.S., Woodson W.R. (1993) Organisation and structure of the 1-aminocyclopropane-1-carboxylate oxidase gene family from Petunia hybrida. Plant Molecular Biology, 23: 1151-1164
Tempe J., Casse-Delbart, F. (1987) Plant gene vectors and genetic transformation: Agrobacterium Ri plasmids. In: Vasil I.K (ed) Cell culture and somatic cell genetics of plants, vol. 6, pp. 25-49
Tepfer, D. (1990) Genetic transformation using Agrobacterium rhizogenes. Physiologia Plantarum, 79: 140-146
Terryn, N., Rouze, P. (2000) The sense of naturally transcribed antisense RNAs in plants. Trends in Plant Science, 5: 394-396
Thayer, S.S., Choe, H.T., Tang, A., Huffacker, R.C. (1987) Protein turnover during senescence. In: W.W. Thomson, E.A. Nothnagel, and R.C. Huffaker (eds) Plant senescence: Its biochemistry and physiology, pp 71-80. American Society of Plant Physiology, Rockville, Md
Theologis, A. (1992) One rotten apple spoils the whole bushel: The role of ethylene in fruit ripening. Cell, 70: 181-184
224
Thompson, K.F. (1957) Self-incompatibility in marrow-stem Kale, Brassica oleracea var acephala. I. Demonstration of a sporophytic system. Journal of Genetics, 55: 45-60
Tian, M.S., Downs, C.G., Lill, R.E., King, G.A. (1994). A role for ethylene in the yellowing of broccoli after harvest. Journal of the American Society of Horticultural Science, 119: 276-281
U, N. (1935) Genome analysis in Brassica with special reference to the experimental formation of B. napus and peculiar mode of fertilisation. Japanese Journal of Botany, 7: 389-452
Van der Straeten, D., Van Wiemeersch, L., Goodman, H.M., Montagu, M. (1990) Cloning and sequence of two different cDNAs encoding 1- aminocyclopropane-1-carboxylate synthase. Proceedings of the National Academy of Sciences USA, 87: 4859-4863
Verma, S.C., Rees, H. (1974) Nuclear DNA and the evolution of allotetraploid Brassicae. Heredity, 33: 61-68
Vriezen, W.H., VanRijn, C.P.E., Voesenek, L.A.C.J., Mariani, C. (1997) A homolog of the Arabidopsis thaliana ERS gene is actively regulated in Rumex palustris upon flooding. Plant Journal, 11: 1265-1271
Wang, C.Y. (1977). Effect of aminoethoxy analog of rhizobitoxine and sodium benzoate on senescence of broccoli. Hortscience, 12: 54-56
Weisz, R., Saunders, M., Smilowitz, Z., Huang, H.W., Christ, B. (1994) Knowledge-based reasoning in integrated resistance management - the colorado potato beetle (Coleoptera, Chrysomelidae). Journal of Economic Entomology, 87: 1384-1399
White, F.F., Taylor, B.H., Huffman, G.A., Gordon, M.P., Nester, E.W. (1985) Molecular genetic analysis of the transferred DNA regions of the root-inducing plasmid of Agrobacterium rhizogenes. Journal of Bacteriology, 164: 33-44
Wilkinson, J., Lanahan, M., Yen, H.C., Giovannoni, J., Klee, H. (1995) An ethylene-inducible component of signal transduction encoded by Never-ripe. Science, 270: 1807-1809
Woodson, W.R. (1987) Changes in protein and mRNA populations during the senescence of carnation petals. Physiologia Plantarum, 71: 495-502
Woodson, W.R., Park, K.Y., Drory, A., Larsen, P.B., Wang, H. (1992) Expression of ethylene biosynthetic-pathway transcripts in senescing carnation flowers. Plant Physiology, 99: 526-532
Yamamoto, M., Miki, T., Ishiki, Y., Fujinami, K., Yanagisawa, Y., Nakagawa, H., Ogura, N., Hirabayashi, T., Sato, T. (1995) The synthesis of ethylene in
225
melon fruit during the early-stage of ripening. Plant and Cell Physiology, 36: 591-596
Yang, S.F., Hoffman, N.E. (1984). Ethylene biosynthesis and its regulation in higher plants. Annual Review of Plant Physiology, 35: 155-189
Yang, S.F., Oetiker, J.H. (1998) Molecular biology of ethylene biosynthesis and its application in horticulture. Journal of Japanese Society of Horticultural Science, 67: 1209-1214
Zambyski, P., Joos, H., Genetello, C., Leemans, M., Van Montagu, M., Schell, J. (1983) Ti plasmid vector for the introduction of DNA into plant cells without alteration of their normal regeneration capacity. The EMBO Journal, 2: 2143- 2150
Zambryski, P., Tempe, J., Schell, J. (1989) Transfer and function of T-DNA genes from Agrobacterium Ti and Ri plasmids in plants. Cell, 56: 193-201
Zarembinski, T.I., Theologis, A. (1994) Ethylene biosynthesis and action - a case of conservation. Plant Molecular Biology, 26: 1579-1597
Zheng, X.Y., Wolff, D.W. (2000) Ethylene production, shelf-life and evidence of RFLP polymorphisms linked to ethylene genes in melon (Cucumis melo L.). Theoretical And Applied Genetics, 101: 613-624
Zhuang, H., Hildebrand, D.F., Barth, M.M. (1995). Senescence of broccoli buds is related to changes in lipid peroxidation. Journal of Agricultural and Food Chemistry, 43: 2585-2591